It has long been labeled the sport of kings, but in this country some have come to regard it as the king of sports.1 Horse racing is now big business in which the state takes considerable interest. Not all involved are in it because of a love of horses. There have been problems generated because of the large sums of money involved in horse race gambling. Cheating and fixing followed gambling. It was not until in the 1960s that a drug problem began to come into prominence. In 1968 the winner of the Kentucky Derby was disqualified after a postrace urine test of the horse, not the jockey, revealed the illegal use of a drug. Now the Illinois Racing Board (“Board”), which closely regulates horse racing on nine tracks, has perceived a new problem, the possible use and abuse of certain substances by the people directly involved in racing licensed as jockeys, drivers, outriders, parade marshals, starters, and assistant starters. The Board believes that drug use risks the safety of the people involved, impairs the state’s financial interest in gambling proceeds, and causes the public to doubt the integrity of racing. Several members of the Jockeys’ Guild (“Guild”) in 1984 asked for Board assistance with the potential drug problem. Representatives of the Illinois Harness Horseman’s Association were also consulted by the Board. The Board subsequently developed its own war on drugs by adopting a comprehensive drug rule which is now at issue.
The Board’s antidrug program offended plaintiffs, Board-approved licensees, who brought this class action in behalf of themselves and other similar licensees, claiming that the urine-testing rule of the Board violated the fourth amendment’s prohibition against unreasonable searches and seizures. Judge Shadur agreed and preliminarily enjoined the Board from enforcing its random urine testing. Judge Shadur also preliminarily enjoined the Board from conducting mandatory urine tests of the plaintiff class without specific articulable facts that meet an objective standard of reasonable cause for believing such individual is under the influence of or has used a *666controlled substance on the grounds of a race track. This latter injunctive requirement largely tracked the regulations to ensure that the standards would be followed and not based on mere “hunch” as plaintiffs feared. Further defendants were enjoined from initially conditioning licensing upon submission to the urine tests prohibited in the preliminary injunction. As laudable as the motivation and intentions of the Board may be in the war on drugs, they are not enough to make up for the constitutional shortcomings of the Board rule. We agree with Judge Shadur, 721 F.Supp. 958, and affirm.
I. HORSE RACING
Horse racing is one of the most ancient of sports. It is believed that chariots drawn by a four-horse team raced in the Olympic games of 700-40 B.C. Riders on single mounts also competed.2 The chariot drivers were the forerunners of today’s harness drivers who now compete in lightweight sulkies drawn by standardbred horses.3 The Olympic bareback riders were the forerunners of today’s jockeys now mounted on the backs of thoroughbreds competing at the run, a faster speed than the harness horse attains at a pace or a trot.4 In the reign of Richard the LionHearted in England between 1189 and 1199, the first racing purse was offered. The first horse-racing trophy on this side of the Atlantic, however, was not awarded until 1665 in the colony of New York.
Although horse racing probably came to North America when Cortez invaded Mexico in 1519,5 the advent of harness racing in this country came later. Beginning with neighbors racing four-wheel buggies on local roads, harness racing developed in the 1800s into the use of the light, two-wheeled sulkies on race tracks. The origin of this type of harness racing with the use of the distinct standardbred horse is an American creation. Although related, the thoroughbred and the standardbred horses are recognized as distinct breeds.
Plaintiffs’ contributions to modern horse racing under their respective licenses must be understood. Jockeys riding thoroughbreds and drivers in sulkies behind stan-dardbred harness horses are recognized, but the duties of the other licensees are less understood.
The outrider in thoroughbred racing is mounted and leads the horses from the paddock area, past the reviewing stands, and into the starting gate area for each race. The outrider makes sure the horses are led out in an orderly fashion without incident. If a horse throws its rider, as sometimes happens, and is running loose, the outrider is responsible for retrieving the horse. The term starting gate does not give an adequate picture of the structure used for starting a thoroughbred race. It is not one large gate stretching across the track. It is more like a series of connected, narrow short stalls, not much bigger than the horse which is led into it. When the horses are all in the gate and ready the electronically controlled door on each individual stall at the command of the starter opens simultaneously, and the race is on.
*667The parade marshal in harness racing performs much the same duties as the outrider in leading the horses to the starting positions, but the start of a harness race obviously cannot be done by the use of the stationary thoroughbred starting gate. A harness race is begun by the use of a moving pace car with retractable bars extending sideways from the back of the car. The starter rides elevated in the rear of the pace car facing backwards where he can observe the moving horses. The harness horses stride into place side-by-side behind the moving pace car and its extended bars. When moving at the proper gait and under control, the pace car barriers are retracted along the side of the car by the starter and the pace car speeds out of the way to permit the horses to move out in competition.
In a thoroughbred race there are a starter and assistant starters with coordinating but different duties. The starter is located just outside the infield rail and elevated where he can observe the horses in the starting gate. He is responsible for ensuring that the horses are properly loaded into their respective sections of the starting gate. He observes their behavior to determine when a fair start is possible. When the horses are all in proper position he releases the starting gate. The assistant starters, however, are afoot on the track with the responsibility of leading the individual horses into their respective gate sections and to help the jockeys maintain control. An assistant starter may enter into the gate and stand on a small side ledge where he can help make sure the horse is under control with head up and facing down the track. When the gate is opened the assistant starter lets go of the horse turning it loose to run, raising his own arms in the air. This shows there was no delay in a starter’s release of a horse possibly causing an unfair start.
In many countries, attendance at horse racing is among the highest of any sport.6 This popularity forecast the intervention of state governments to protect a source of state revenue from wagers on horse racing totalling above one billion dollars per year in Illinois, and to protect the integrity of horse racing.7 So it is that the Board sought to do its duty under its broad legislative mandate, the Illinois Horse Racing Act of 1975 (“Act”),8 to regulate and control horse racing.9
II. THE BOARD RULE
The Act gives the Board discretion to issue occupational licenses to persons working on Illinois race track grounds privately owned by race track organizations. Licenses may be refused for a variety of things, from criminal convictions to a violation of the rules and regulations of the Board.10 So also may a license be suspended or revoked by the Board for just cause, including for violation of any Board rule or regulation.
The attendant power is given to the Board to promulgate reasonable rules and regulations for the purpose of administering the Act and to prescribe reasonable rules and regulations under which Illinois horse race meetings are conducted. It is further statutorily provided that this rule-making power includes the general power to adopt rules and regulations to prevent practices determined to be detrimental to the public interest, to provide for the best interests of horse racing, and to impose penalties for violations.11 Civil penalties may range up to $5000 against individuals and up to $10,000 against organizations.12 The Board’s statutory power also includes the power to eject and exclude from the *668race grounds any licensee whose conduct or reputation is such that his presence may, in the opinion of the Board, call into question the honesty and integrity of horse racing or interfere with the orderly conduct of horse racing.13 There can be no doubt that Illinois horse racing is a highly regulated business in all respects, both for thoroughbred and standardbred horse racing, extending from the breeding of the horses 14 to the care of chronic horse race gamblers.15
The Board proceeded to adopt the challenged substance abuse rule (“Rule”) and guidelines for its implementation (“Guidelines”), but the Board’s broad statutory authority is not what is challenged. It is explained in Rule § 508.10(b) that the Rule’s purpose is “to prevent practices in horse racing that are detrimental to the public interest, to promote the best interests of horse racing, and to cooperate in the establishment of a national substance abuse rule for racing as proposed by the National Association of State Racing Commissions.” In particular the Board explains in Rule § 508.10(c) that it is concerned with drug use in the racing industry because of the impact on an individual licensee’s ability to perform his or her duties, and because addiction may make that licensee peculiarly susceptible to bribes or other improper influences.16
Rule § 508.50(a) proceeds to prohibit the use on the grounds of any race track any controlled substance or any prescription drug unless such substance was obtained directly, or pursuant to a valid prescription or order, from a licensed physician. That rule prohibition is specifically made applicable to jockeys, drivers, starters, assistant starters and outriders. Guidelines at 2. To implement that prohibition Rule §§ 508.-50 & 508.802 provide for two kinds of urine testing: testing based on individualized suspicion and testing entirely at random. Judge Shadur considered random testing to be more problematic and therefore considered it first, as do we. We track Judge Shadur’s findings.
A. Random Testing
The Board determines the volume and frequency of random testing at each race meet, as well as which selected racing programs are subject to testing. Guidelines at 2. The names of all licensees who appear as participants on the official program are placed in a locked container secured by the race stewards, the stewards being the on-site state supervisors of the racing meet.17 A steward then may draw up to five names of licensees from the container, with a representative of the Guild, the Illinois Horseman’s Benevolent and Protective Association, or the Illinois Harness Horseman’s Association having been invited to witness the selection process. Rule § 508.80(c). Following the drawing the stewards locate and notify the licensees selected for random testing. Each licensee drawn must report to the designated sample collection area and provide a urine sample to the stewards or their designee before the last race on that day’s racing program. Rule § 508.80(d); Guidelines at 2.
No licensee is required to provide a sample more than three times in a race meet. Rule § 508.80(e); Guidelines at 2. Judge Shadur found that the Board has now agreed, though it does not appear in the Rule or Guidelines, that a licensee will not be randomly tested more than five times per year. If a licensee’s name is drawn more than three times in any meet (or presumably more than five times in any year), the stewards will disregard the selection, return the name to the container, and draw another name. Rule § 508.80(e); Guidelines at 2-3. To implement that limitation the stewards maintain a confidential log of all licensees selected for random testing. Id. at 3.
*669B. Individualized-Suspicion Testing
Any particular licensee who is suspected to be in violation of the Rule will receive written notice from the stewards stating that he or she will be tested and giving the reason justifying the testing. Id. at 3. Individualized-suspicion testing must be based on a finding that the licensee is under the influence of drugs as explained in the Guidelines where it is provided that:
Such a finding by the stewards shall be based upon corroborated evidence or observable phenomena, such as direct observation of drug use or possession and/or physical symptoms of being under the influence of alcohol or drugs such as erratic behavior. These are examples of conduct which would satisfy individualized suspicion standards, however, this list is noninclusive.
Id. at 3.
C. Urine Collection and Testing Procedures
Any licensee selected for urine testing, either at random or because of individualized suspicion, must present himself or herself at the designated collection site. The site is secured by Board representatives. No unauthorized personnel are permitted in any part of the site where urine specimens are collected or stored. Id. at 4. The routine is that before providing the sample the licensee must present personal identification. The sample may be provided in the privacy of a stall or other partitioned area that allows for individual privacy. Before entering the stall the licensee must remove all unnecessary outer garments and leave personal belongings outside. Bluing agents may be added to the toilet water to prevent adulteration. Id. at 4.
Upon receiving the sample from the licensee, the Board representative determines whether the sample contains a minimum of fifty milliliters of urine (one-half of the bottle). If not, the licensee may return later in the day to provide a sufficient sample, but it must be before the last race that day. If unable to provide a sample, the licensee may be requested to return the next day. Id. at 4. Once the Board representative receives the sample it must be checked for signs of tampering, such as obvious temperature or color difference. Then the representative separates the sample into two parts, a “laboratory” sample and a “referee” sample. Only the laboratory sample is tested initially. Rule § 508.50(d); Guidelines at 6. Both sample bottles are then sealed and tagged. The Board maintains stringent procedures for assuring the chain of custody of the sample. The licensee is required to verify that he or she has witnessed the identification, splitting, and sealing procedures by signing a form provided by the Board. Rule § 508.50(c); Guidelines at 5-6. All sealed urine samples are kept in a locked refrigerator pending pickup by the laboratory. Id. at 7.
After analyzing the “laboratory” sample for possible adulteration, the laboratory screens the urine using an immunoassay test that meets FDA requirements for commercial distribution. Id. at 8. Immunoassay testing uses antibodies to detect the presence of similar drugs and their metabolites in the urine. In this instance the test screens for “Controlled Substances,” defined as any substance listed in 21 U.S.C. § 812 (not including later amendments to that statute). Rule § 508.20; see also id. § 508.50(a). When the concentration level of any of the following substances meets the indicated cutoff level, the test is considered “positive”:
Marijuana metabolites 100 ng/ml18
Cocaine metabolites 300 ng/ml
Opiates 300 ng/ml
Amphetamines 1000 ng/ml
Barbiturates 300 ng/ml
Guidelines at 8. Except for marijuana metabolites, those test levels are subject to change as advances in technology permit identification and quantification at lower concentrations. As for marijuana metabolites, the level is set at 100 ng/ml to reduce the chance that a licensee exposed to passive inhalation will produce a “positive” reading.
*670If the initial screening reflects any positive reading, a Gas Chromatography/Mass Spectrometry (“GCMS”) test is used to confirm the results. GCMS isolates an individual drug or metabolite and identifies it by a characteristic fragmentation pattern similar to a fingerprint. If the GCMS test confirms the sample as positive, both the resealed laboratory sample and the unsealed referee sample are retained by the laboratory in long-term frozen storage for at least one year. Id. at 10. If the test yields a negative result, both the laboratory and referee samples are kept in refrigerated storage for seven days and then destroyed.
Any laboratory selected by the Board for conducting the testing must have a comprehensive quality assurance program and must also participate in certification and accreditation programs. The Board also employs a Drug Testing Assessment Program to monitor the laboratory’s performance through the use of blind quality controls. Id. All test results are treated as confidential. They may be used solely with respect to a ruling issued pursuant to the Rule or in any administrative or judicial hearing regarding such a ruling. All such information is exempt from public disclosure under the Illinois Freedom of Information Act.19 Guidelines at 12. If a penalty is warranted the Board must consider the offender’s history of Rule violations, age, experience, and the potential of the conduct for physical harm to human and equine participants at the race meeting. Rule § 508.60(a). Increasing penalties are provided for each new positive test. Rule §§ 508.60(a)-(c); Guidelines at 13-14.20 Any licensee who fails to submit to a urine test when requested to do so must be suspended. Rule § 508.50(b).
It is the licensee’s responsibility to give advance written notice that he or she is permissibly using a controlled substance or prescription drug. Rule § 508.50(a). Also any person contesting the results of a positive test may request a hearing before the Board. In addition, the licensee may have the referee specimen tested at a laboratory of his or her choice. Guidelines at 14.
In his concluding findings of fact Judge Shadur also found that all of the substances tested for under the Rule can cause impairment in humans for a period of time after use, the extent of impairment depending on a number of factors, including the amount of the drug ingested. Urine testing does not, however, necessarily determine whether the licensee was impaired at the time the urine specimen was provided. Urine tests indicate only whether the metabolites of a drug are present in the system. A positive urine test does show that the licensee has ingested or been exposed to a particular drug at some time in the past. It does not reveal how much of the substance was ingested, when it was ingested, or what its effect may have been on the licensee after it was ingested.
The findings of fact are not seriously attacked and we accept them as not clearly erroneous. They illustrate the care with which the Board sought to implement its drug program. Additional findings will be discussed when relevant to the preliminary injunction issues.
III. ISSUES
What is being reviewed is the appropriate exercise of discretion by Judge Shadur in the grant of the preliminary injunction prohibiting random urine testing under the Rule or conditioning a license upon submission to the enjoined urine tests. The *671Board’s program as to reasonable suspicion testing was also enjoined by Judge Sha-dur’s order but only to insure that the testing be based on specific articulable facts that meet an objective standard of reasonable cause for believing that the person is under the influence, or has used a controlled substance, on the race track grounds.
In issuing the preliminary injunction Judge Shadur was guided by the preliminary injunction discussion of this court in Roland Mach. Co. v. Dresser Indus., Inc., 749 F.2d 380, 386-88 (7th Cir.1984). Since the Board in the district court concentrated its argument on the merits so did the district court. The primary factor at issue was the likelihood of success of the class plaintiffs on the merits of the random testing rule. However, Judge Shadur found that class plaintiffs had no adequate remedy at law and would otherwise suffer irreparable harm without the issuance of the preliminary injunction. It was also found that the balancing of harms weighed in class plaintiffs’ favor and that the issuance of the preliminary injunction would not dis-serve the public interest. The other aspects of the granting of the preliminary injunction are questioned by the Board on appeal, but all generally agree that the case is largely controlled by recent fourth amendment precedents of the Supreme Court.
IV. DISCUSSION
It is necessarily conceded that urine testing of individuals is a fourth amendment “search” and therefore must meet the constitutional standard of reasonableness. Schaill v. Tippecanoe County School Corp., 864 F.2d 1309, 1311-12 (7th Cir.1988); see also Skinner v. Railway Labor Executives’ Ass’n, 489 U.S. 602, 617, 109 S.Ct. 1402, 1412-13, 103 L.Ed.2d 639 (1989). The Board recognizes that an individual has a reasonable expectation of privacy in the ordinarily private act of urination and also in what a chemical analysis of a urine specimen may reveal. However, in an effort to avoid the necessity for a search warrant the Board attempts to bring its rule within a “special government needs” exception in circumstances going beyond the ordinary needs of law enforcement. If the Board drug effort comes within that exception then it may conduct urine testing without a warrant, without meeting a probable cause standard, without individualized suspicion, and without having to show that the group to be tested has or is likely to suffer any incidence of drug abuse. In that regard two recent Supreme Court cases must be examined for guidance.
In National Treasury Employees Union v. Von Raab, 489 U.S. 656, 109 S.Ct. 1384, 103 L.Ed.2d 685 (1989), a five-to-four decision, the Court held that the United States Customs Service could implement its employee drug-testing program. That program required certain of its employees— those applying for promotion to positions directly involving interdiction of illegal drugs, or for promotion to positions requiring employees to carry firearms — to produce without a warrant urine samples to be analyzed for evidence of illegal drug use. The Court approved the use of the program as a reasonable fourth amendment intrusion because special governmental needs, beyond the normal need for law enforcement, were served. It reached that conclusion by balancing the individual’s privacy expectations against any impracticability the government’s interests might suffer in first requiring a warrant or some kind of individualized suspicion in the particular factual context. Id: at 665-66, 109 S.Ct. at 1390-91. The court noted that the drug-testing program did not serve the ordinary needs of law enforcement, and that the tests could not be used in a criminal prosecution of the employee without consent. The program was designed to deter drug use by the particular employees involved in the interdiction of illegal drugs, even by employees who are physically fit with unimpeachable integrity and judgment. See id. at 670, 109 S.Ct. at 1393. It was found that possible drug use by an employee with those drug interdiction responsibilities could corrupt the important government nationwide- antidrug effort. Likewise the public interest, it was found, demands effective measures to prevent, because of *672risk to others, the promotion of employee drug users to positions requiring the carrying of a firearm. Against those public interests the court weighed the interests of the particular employees to be free of interference with their individual liberty and privacy and found the balance to weigh in the government’s favor. Certain forms of public employment were found necessarily to reasonably diminish privacy expectations. Id. at 671-72, 109 S.Ct. at 1393-94. The details of the particular urine-testing procedures were found to be sufficiently effective and not to be an arbitrary and oppressive interference with privacy in the context of the government’s attack on drug abuse, labeled one of the most serious problems confronting our society. The validity of the drug-testing program was also found not to be impugned by the mere circumstance that most of the employees tested were entirely drug free. The strong federal government need to prevent and deter drug use was found to be ample justification for reasonable warrantless searches. Justice Scalia dissented because in his view neither the frequency of drug use nor its connection to possible harm to others had been demonstrated. That is in contrast to his view under different facts in Skinner v. Railway Labor Executives’ Association, a case we next must examine.
Skinner, a companion case to Von Raab, discusses the same constitutional principles but in a different factual context. In Skinner drug regulations were adopted under the Secretary of Transportation’s statutory authority to adopt safety standards for the railroad industry. 489 U.S. at 606, 109 S.Ct. at 1407. These safety standards required, among other things, that railroads see that urine samples for testing are collected from crew members or other employees directly involved in a railroad accident, that is from those employees who are engaged in safety-sensitive tasks. Railroads were also given authority, but not required, to do similar testing of those same types of employees who violate certain safety rules. The stated purpose of the testing rule is to ensure the safety of the traveling public and the employees. Id. at 620-21,109 S.Ct. at 1414-15. Note is taken that railroading is a pervasively regulated industry in order to ensure public safety. Id. at 627, 109 S.Ct. at 1418. The court makes it clear, however, that the bodily security of employees employed in a regulated industry must not always be considered minimal. Id. at 628, 109 S.Ct. at 1418-19. A train is characterized as lethal when operated by persons under the influence of drugs. Because the employees subject to the urine test discharge duties fraught with risk of injury not only to themselves but to others, the government was found to have a compelling interest in drug testing without a showing of individualized suspicion. Id. at 628-29, 109 S.Ct. at 1418-19. The privacy interests of the railroad employees were, therefore, outweighed. Justice Scalia this time joined in the majority opinion. Justice Stevens concurred in part and in the judgment. Justice Marshall dissented joined by Justice Brennan. The dissent may be summed up by the view expressed that the first and worst casualties of the war on drugs will be the precious liberties of citizens. See id. at 636, 109 S.Ct. at 1423 (Marshall, J., dissenting). The dissent would not dispense with the probable cause requirement of the fourth amendment merely because it becomes “impracticable.”
Those two eases, Von Raab and Skinner, consider the same constitutional principles, but in different factual circumstances. It can be seen in comparing the holdings in Von Raab and Skinner with what we know so far about random urine testing in horse racing that the validity of that testing is in jeopardy. We need, however, briefly to go further with the constitutional analysis.
Horse racing is pervasively regulated by the state of Illinois, but that factor alone is not enough. The horseracing regulations are stated in the Board’s Rule and Guidelines to be for the purpose of preventing fraud and practices detrimental to the public interest, to promote the best interests of horse racing, and to cooperate in the establishment of a national substance abuse rule for racing as has been pro*673posed.21 The particular concerns are stated to be the drug impact on an individual licensee’s ability to perform his or her duties and the likelihood that addiction may make an individual more susceptible to bribes or other improper influences.22 Those are the government interests, safety and integrity, to be weighed against the privacy interests of the licensees. Drug-using licensees were argued to be more susceptible to bribes or other improper influences, but the district court found insufficient evidence to support that assumption, only speculation and argument as characterized by Judge Shadur.
The integrity of the race affects, among other things, not only who receives the winning purse but also the market reputation of the particular horse involved. The Board’s integrity concerns' are very important but the Board fails to explain why the criminal remedies for bribery, blackmail, race fixing, and other illegal activity are inadequate. Those ordinary criminal remedies not only punish, but also may deter the conduct purportedly affected by the Rule. Moreover, the sweep of the Rule is overly broad. The integrity risk is greatest with jockeys and drivers who can in various ways directly affect the race outcome. As to the other licensees, the starter might start the race before a particular horse was ready. The assistant starter might hold the horse slightly too long in the starting gate or do something else to distract the horse. The starts, however, are videotaped and the district court found little realistic opportunity for the starter or assistant starters to throw the race at that stage. In any event, that much of the government interest, though important, does not weigh enough when on the scale with the fourth amendment. The Board’s integrity interest does not compare with the much more important public interests illustrated in Skinner and Von Raab.
There is also another aspect of the integrity of racing. If the public comes to believe it is all fixed or controlled, racing could lose much of its fascination. There has always been a degree of public skepticism. The result would be a decline in parimutuel betting, diminishing the state’s share of the take. Shoemaker v. Handel, 795 F.2d 1136, 1142 (3d Cir.1986), cert. denied, 479 U.S. 986, 107 S.Ct. 577, 93 L.Ed.2d 580 (1986), found that to be a sufficient justification, but that case predates Von Raab and Skinner. Those Supreme Court cases, as we read them, leave little leeway for revenue considerations, as important as state revenue is. A decrease in state revenue does not equate with loss of life in a train accident, or the possibly irresponsible and dangerous use of guns by Customs officers or other dereliction of duty with broad impact. Revenue loss does not directly affect public safety and the Board fails to explain why normal law enforcement cannot protect the state revenue. A good cause and a practical solution is not enough to make a constitutionally illegal search legal.
The Board, however, also advances its own safety rationale, again an important consideration, but in the horseracing context the safety concerns are factually much more limited than in Von Raab or Skinner.
There is physical danger in horse racing to the horses and the participants as borne out in the record, but little to the public spectators, although occasionally a horse may go over the rail in an accident. The district court findings note that the Guild has forty permanently disabled members. In 1987 over 100 members suffered racing injuries severe enough to cause at least one week of disability. Nationally about two jockeys a year are killed in racing accidents. The accidents are not shown to have any particular relationship to drug use by participants. Jockeys and drivers are at the greatest risk, but they are at risk in racing regardless of drug use. Drug use no doubt would increase that risk as alert, unimpaired, and physically fit participants are critical for the safety of themselves, others, and the horses.23
*674Many accidents are caused by some natural weakness in the horses. Horses themselves may be at the greatest physical peril, which in turn endangers the participants.24 A thoroughbred weighing over 1000 pounds running at about forty-five miles per hour on small and fragile legs on hard or muddy tracks side-by-side with other horses sometimes bumping, obviously creates risks. A jockey obviously has great influence on his horse, and could possibly cause it to get off stride, to falter and fall. The signals a jockey gives his horse for speed and direction can be obvious when the whip is used, but subtle when only leg pressure or weight shifting is applied. A jockey impaired by drugs would no doubt greatly increase these risks for himself, his horse, and the other participants.
Harness drivers have similar safety problems. Sitting low'behind their horses the vision of the drivers is restricted. Drivers have to maintain a safe interval between sulkies both front and rear and side-to-side. A sulky wreck can involve others in a dangerous mix of horses, people, and equipment. It cannot be said that participants impaired by the possible use of drugs is the principal cause of those risks, but drug use would obviously exacerbate them.
Urine testing, as the district court found, has limited use for these purposes as it does not measure the licensees’ present impairment, but only reveals that drugs were previously ingested and were in the person’s system. The biggest value of urinalysis might prove to be deterrence, but that is not an unimportant consequence. See Skinner, 489 U.S. at 628-30, 109 S.Ct. at 1418-20.
The other licensees are at less physical risk, outriders, parade marshals, starters, and assistant starters.25 The district court found this was minimal risk to some and no “real safety risk” for assistant starters, for instance. He found the safety rationale did not extend to all licensees.26
After a review of the record we have a more serious view than does the district court as to the safety rationale in regard to the other licensees not actually in command of a horse, but we do not undertake to make our own factual findings. The enhanced risk we see for other licensees besides the jockey and drivers would not change the constitutional analysis or result reached by the district court. We also see some risk from any impaired licensee working as part of the racing team with related responsibilities which could affect the safety of other licensees. A drug-impaired starter, for instance, although safely outside the track, could start a race when a horse was not ready and not only affect the outcome of the race but cause injury to the jockey or the assistant starter in the gate as well as the horse. All these risks, however, are largely confined to the track and the licensees, not to the public spectators, which is in some contrast to motor racing.
Another Board argument is that the privacy interests advanced by plaintiffs are already reduced because horse racing is *675pervasively regulated. That is a view expressed in Shoemaker v. Handel, where the Third Circuit held that the intensive state regulation of horse racing reduces the justifiable privacy expectations of the licensees. When participants chose to be a part of racing by accepting a state license, they knew what to expect. 795 F.2d at 1142. We see some merit in the Board’s argument but the Supreme Court has rejected the notion that “the interest in bodily security enjoyed by those employed in a regulated industry must always be considered minimal.” Skinner, 489 U.S. at 628, 109 S.Ct. at 1418-19. We do not find that pervasive regulation of horse racing’s other aspects is enough to reduce the licensees’ justifiable privacy expectations and validate otherwise unconstitutional individual urine testing.
In Schaill this court approved random drug testing for interscholastic high school athletics. 864 F.2d at 1809. This case was also before the Supreme Court decided Von Raab and Skinner. In Schaill the school context was strongly taken into account with the needs for swift and informal disciplinary measures. The constitutional issues are well-analyzed, but the factual circumstances are sufficiently different so as not to control the present case. As Judge Shadur noted as to Schaill, this court mentioned that athletics was an extracurricular activity and that the testing would not result in loss of employment, criminal penalties, academic penalties or even suspension or expulsion. Id. at 1319. That is not entirely true in the horseracing context.
The Board also refers us to Taylor v. O’Grady, 888 F.2d 1189 (7th Cir.1989), which considered the urine testing of prison personnel which had been enjoined by the district court as a violation of the fourth amendment. This court affirmed in part and held the regulation to be too broad as governmental interest in the integrity of its work force was not sufficiently strong to overcome the privacy interests of all the employees. But this court separated out of the broad prison testing program those personnel with direct contact with the prisoners. The governmental interests were high because the presence of drugs in prisons, possibly sold on the grounds, constituted a major security risk, possibly leading to violence between prisoners and armed guards and even escapes. Horse-racing circumstances, however, cannot equate with the potential dangers of drugs in a penitentiary-
In Serpas v. Schmidt, 827 F.2d 28 (7th Cir.1987), cert. denied, 485 U.S. 904, 108 S.Ct. 1075, 99 L.Ed.2d 234 (1988), this court considered fourth amendment objections to a different segment of horse racing, those licensed people who help care for and exercise horses in between races, called “back-stretchers.” The Board purported to have the authority to conduct warrantless searches of the on-track dormitories of those licensees and to make investigatory stops and seizures of the licensees within the track enclosure. The Board claimed there was, among other things, implied consent to the warrantless searches when licensed employment was accepted. The Board policy was designed to protect horses from being administered drugs or to protect against the application of a mechanical device to the horse. The device is referred to as a buzzer and may affect the speed of the horse and possibly the outcome of the race. The court did not doubt the threat posed by the drugs and buzzers but it was held not enough to avoid the fourth amendment. Judge Eschbach dissented on other grounds.27
V. CONCLUSION
We do not see a sufficient justification for the relaxation of constitutional safeguards in the horse-racing context, particularly after the tightening up of fourth amendment exceptions under Skinner and Von Raab. There are good and valid reasons to strive to keep drugs out of horse racing, but, as important as that is, that cannot be accomplished by disregarding the Constitution. The Board along with all well-intentioned licensees, assuming the drug threat is as serious as the Board believes it to be, should be able to find a *676constitutional way to accomplish an effective drug program. If not, horse racing may not long survive as the king of sports as it is viewed by some. It may even deteriorate into no sport at all.
Affirmed.
. Some animal activists and horse racing critics, however, refer to the king of sports as a "ticking time bomb” in need of reform. See Maggitti, Don’t They?, Animals' Aghnda, Nov. 1990, at 19, 19.
. Only recently has there been reason to believe that horses were ridden as early as 4000 B.C., about 2500 years earlier than previously believed, and 1000 years before the invention of the wheel. Hill, Where There Were Fires, There’s Still Smoke, MHQ: The Quarterly Journal of Military History, Winter 1991, at 74, 75.
. For an overview of horse racing see 6 New Encyclopaedia Britannica 68 (15th ed. 1987) (Mi-cropaedia Ready Reference) [hereinafter Britannica]. It is here that it is mentioned that horse racing has become known as the king of sports, but horse racing may be a loser to baseball, the commonly accepted "national pastime." See Flood v. Kuhn, 407 U.S. 258, 264, 92 S.Ct. 2099, 2103, 32 L.Ed.2d 728 (1972).
For a more comprehensive view of harness racing see T. Ainslie, Ainslie’s New Complete Guide to Harness Racing (1980), and J. Hervey, The American Trotter (1947).
For a more comprehensive view of thoroughbred racing see B. Gifford, A Day at the Races (1988).
. The pace is a more artificial gait than a trot. A pace can be recognized when the horse moves both legs on the same side of its body in unison, whereas at the trot the horse moves in unison the two diagonally related legs, as the right front and left hind, or vice versa.
. J. Hervey, supra note 3, at 16.
. Britannica, supra note 3, at 68.
. State horse-racing regulations in this country, even though pervasive, are in sharp contrast with horse racing in the Soviet Union, where, at least up until the present time, the government not only owned the track but all those involved were government employees, including the jockeys. Id. at 70.
. Ii.i..Ri:v.Stat. ch. 8, paras. 37-1 to -53.
. See id. para. 37-15(a).
. Id. para. 37-15(c).
. Id. para. 37-9(b).
. Id. para. 21-9(1).
. Id. para. 37-9(e).
. Id. para. 37-33.1.
. Id. para. 37-31.1.
. Rule §§ 508.30-.40 outline the procedures for testing for alcohol, but the class plaintiffs do not challenge that aspect of the Rule.
. Board rules give the stewards nearly absolute control over race track activities subject to Board review.
. Nanograms per milliliter.
. Ii.l.Rev.Stat. ch. 116, paras. 201-211.
. For the first violation, the licensee is to be suspended or assessed a civil penalty of up to $1000.
For a second violation, Board must suspend the licensee pending the completion of a substance abuse treatment program licensed by the Illinois Department of Alcoholism and Substance Abuse, or a state-licensed treatment program in another state. Upon completion of the program the licensee is subject to follow-up urine testing for one year, but not to exceed four tests in a year. Follow-up urine testing is a condition of “re-licensure.” Any individual suspended for a second violation is entitled to a hearing.
For a third violation, the offender's license must be revoked. If an individual’s license is revoked, he or she is entitled to a hearing.
. Rule § 508.10(b).
. Rule § 508.10(c).
.In 1985 the Board instituted a pilot drug-screening program of jockeys and harness drivers in order to learn the scope of the problem. *674Out of a total of 568 harness drivers tested, seventy were found positive for cocaine, THC, or both, a positive rate of 12%. Of 162 jockeys tested, 52 were found positive for cocaine, THC, or both, a positive rate of 32%. Plaintiffs attack the validity of the testing methods.
A drug problem to some degree must exist, as the Guild has a drug abuse counseling program.
. The press reported three thoroughbred horses died in the 1990 Breeders Cup races at Belmont Park; Go For Wand, Mr. Nickerson, and Shaker Knit collapsed for physical reasons. Fragility Puts Horses’ Lives in Peril, Chi.Trib., Oct. 30, 1990, § 4, at 4, col. 1.
. Bob DuPont, a former assistant starter, vivid ly recounts; "A lot of horses have to be ... held by the tail because they’d flip over backward in the gate. Sometimes you wind up down inside underneath that horse. He’ll pull you down in there and you better get out fast.” B. Gifford, supra note 3, at 131. In Mr. DuPont’s view ”[w]orking in the starting gate is a bad place.” Id. at 130. He also describes what can happen in the gate to the jockey: "[I]f a horse rears, the rider can fall off behind him and be squeezed in the gate, which is locked, cam locked, and he can easily be killed, crushed to death.” Id. at 131.
.Even outriders can have risk as related by Ben Larraine, an outrider: "A real problem is when the rider gets in trouble. He breaks a rein or stirrup, gets hung up in the stirrup and you go after him. That's when you put your neck on the line.” Id. at 98.
. Four members of the court dissented from the denial of rehearing en banc.