Keene Corporation appeals from a final judgment of the United States District Courts for the Eastern and Southern Districts of New York (Charles P. Sifton, Judge) awarding plaintiff Roberta Kranz, as the executrix of the estate of Lee Lewis, $226,038.49 for personal injury, wrongful death, and loss of consortium. In re Joint E. & S. Dists. Asbestos Litig., 798 F.Supp. 925 & 798 F.Supp. 940 (E. & S.D.N.Y.1992). The claims arose from Lewis’s exposure to asbestos products manufactured by Keene’s subsidiary, the Baldwin-Ehre-UHill Company (“BEH”). For the reasons stated below, we reverse and remand for a new trial.
*348BACKGROUND
The Explosion Of Asbestos Litigation
One of the greatest challenges facing both state and federal courts is the crush of tort suits arising from the extensive use of asbestos as flame-retardant insulation throughout much of this century. Asbestos litigation today constitutes the largest mass toxic tort in the United States. See In re Joint E. & S. Dists. Asbestos Litig., 125 F.R.D. 60, 63 (E.D.N.Y.1989) (hereinafter “Drago”). To date, more than 200,000 asbestos cases have been filed by injured persons and their heirs, and as many as 250,000 additional cases may be filed in years to come. See Stefan Fatsis, Fallout from Asbestos Crises Still Clogs Court, Defies Solutions, L.A. Times, August 2, 1992 at Al.
Asbestos fibers — which contain highly toxic carcinogens — have debilitated and killed many workers. Asbestos-related injuries are characterized by prolonged latency periods and typically do not become apparent until years, and in some cases decades, after exposure. Thus, many potential plaintiffs were unable to recover from the manufacturers and distributors of such products because of statutes of limitations that started running from exposure to asbestos rather than from manifestation of disease.
Recognizing the inequity, New York amended its statute of limitations in 1986 to provide some recourse to asbestos victims. The new legislation triggered the limitations period from discovery of the disease, New York Toxic Tort Reform Act of 1986, L.1986 ch. 682, § 2 (codified at N.Y.C.P.L.R. § 214-c (McKinney 1990)), and explicitly revived previously time-barred asbestos actions. L.1986, ch. 682, § 4, reprinted after N.Y.C.P.L.R. § 214-c (McKinney 1990).
Because of these changes, both the state and federal courts were swamped with asbestos suits. In response, the federal courts tried several innovative procedures. For example, recognizing that “[t]he heyday of individual adjudication of asbestos mass tort lawsuits has long passed,” the Judicial Panel on Multidistrict Litigation ordered the pre-trial consolidation of 26,639 pending asbestos cases in July 1991. In re Asbestos Prods. Liab. Litig. (No. VI), 771 F.Supp. 415, 419 (J.P.M.L.1991).
In New York, the Chief Judges of the Second Circuit, the Southern District, and the Eastern District transferred all cases filed in either district to the district judge in this action for purposes of discovery. 798 F.Supp. at 944. We commend Judge Sifton for his masterful stewardship of these cases. Eventually, the cases approached the Rubicon of either settling or going to trial. To facilitate settlements and provide for manageable trials, the cases were “subdivided by the location in which the plaintiff suffered primary exposure.” Id.
The Consolidation Here
In the instant action, 600 cases were consolidated. The thread upon which all 600 cases hung was that each plaintiff had been exposed to asbestos in one or more of over 40 power-generating stations, or “powerhouses” as they are called, in New York State.
Forty-eight were selected from the 600 cases for trial on a reverse-bifurcated basis, i.e. damages to be tried first and then liability. The damages trial began on April 1, 1991. Each of the 48 plaintiffs had named as defendants between 14 and 42 manufacturers or distributors of asbestos-containing products. Of these, 25 appeared at trial as direct defendants. Several of the defendants im-pleaded third-party defendants. For example, on March 18, 1991, 13 days before the trial began, Judge Sifton allowed defendant Owens-Corning Fiberglas Corporation to im-plead over 200 companies. Some of the third-party defendants, in turn, impleaded fourth-party defendants.
During the four-month damages trial, evidence of the debilitating diseases and/or deaths of all 48 plaintiffs was presented to the jury. Often, the plaintiffs themselves would testify to the devastating consequences suffered as a result of asbestos-related disease. Where, as in Kranz-Lewis’s case, a particular victim had died prior to trial, evidence regarding his disease and death was presented by family members. A parade of medical doctors testified on the etiologies and pathologies of the asbestos-related diseases suffered by each of the plaintiffs. Econo*349mists testified concerning the present value of past and future income streams, and the dollar value of ordinary household services.
In addition, detailed testimony for each victim was necessary concerning his. degree of impairment, specific medical history, emotional state, and medical .prognosis. Further complicating matters, the jury had to sift through each victim’s medical history to determine whether factors other than asbestos, such as smoking, were responsible, in whole or part, for his physical complaints. For example, Mr. Lewis’s son testified that his father smoked cigarettes until the mid-1950’s, suffered episodes of chronic coughing and hoarseness, switched to pipes and cigars, switched back to cigarettes, experienced a scare after an episode of hoarseness, and ultimately quit after developing a polyp' on his voice box in the early 1960’s. Claims by spouses and children' presented extensive plaintiff-specific evidence.
After four months of such evidence, the jury returned verdicts for 45 of the plaintiffs for an aggregate of over $94 million. Kranz-Lewis’s damages were calculated as $1,682,-795, including $1,250,000 for “Pain, Suffering and Other Non-Economic Losses to Decedent.”
The liability portion of the trial began on September 11, 1991. During this phase, the jury was presented with a dizzying amount of evidence regarding each victim’s work history. Where a victim, like Lewis, had died before trial, the sites where he had worked during his career, the types of asbestos-containing products with which he had been involved, and the identity of the manufacturers or distributors of the asbestos products to which he may have been exposed were reconstructed through the testimony of family members and co-workers.
The testimony of just one plaintiff illustrates the cosmic sweep of the factual data that the jury had to absorb. That plaintiff, Hubert Feeley, testified that from 1953 until 1974, he worked in “hundreds of’ buildings; “could not keep track of all of them;” used “[p]ipe covering, block cement, asbestos cloth, all different sorts of cement;” worked for “[tjwenty-five or so” employers; travelled as an asbestos worker to Alaska, Egypt, Wyoming, Minnesota, West Virginia, Connecticut, White Plains and New Jersey; worked in “[ojffice buildings, high-rises, shopping centers, [and] state office buildings” including the Chase Manhattan Building, the Exxon Building and the Holiday Inn; and worked in at least seven powerhouses throughout the greater New'York area. He also testified that he used at least ten different products while working for one of his many employers. ■ He candidly testified on direct examination that “of-the hundreds of buildings [he ‘ had] worked on,” “[m]aybe eight or nine were powerhouses.” Finally, the longest period that he could recall working at any one powerhouse was “about six months.” (It should be remembered that the common thread supporting consolidation of all these cases was asbestos-exposure in a powerhouse.).
After three months, plaintiffs rested on December 4, 1991. For the next three months, the defendants presented their case. The district court and the,lawyers valiantly attempted to maintain the identity of each claim throughout the trial; The jury was instructed on several occasions to consider each case separately and each juror was given a notebook for this purpose. Thanks to the effective settlement techniques of the district judge and a special master, only two plaintiffs remained by the time the jury rendered its liability verdict. It concluded that appellant Keene Corporation was 9% liable for the Kranz-Lewis damages.
Following the verdict, Keene moved for judgment as a matter of law, a new trial, or other post-verdict relief, contending, inter alia, that the district court’s decision to consolidate the 48 cases for trial constituted prejudicial error. The district judge rejected this argument without extended discussion; and, after molding the verdict in accordance with various New York statutes to add interest and to reflect different degrees of fault among defendants, entered a judgment for Kranz against Keene for $226,038.29.
DISCUSSION
Addressing the complaints of hundreds of thousands of severely injured asbestos plain*350tiffs, while safeguarding the rights of the defendants, all the while searching for equitable resolutions in each case, is a herculean task. Many of the asbestos victims suffered exposure for decades and at many different worksites. Finding an appropriate forum to resolve all these claims with minimal delay is the goal. Faced with this challenge, district judges throughout the country have reacted with commendable ingenuity. Pre-trial consolidation for the purposes of discovery, the appointment of special masters to expedite settlement, and, especially, the liberal use of consolidated trials have ameliorated what might otherwise be a sclerotic backlog of cases.
Federal Rule of Civil Procedure 42(a) provides that “[w]hen actions involving a common question of law or fact are pending before the court, it may order a joint hearing or trial of any or all the matters in issue in the action....” As we recently noted,
Consolidation of tort actions sharing common questions of law and fact is commonplace. This is true of asbestos-related personal injury cases as well.
The trial court has broad discretion to determine whether consolidation is appropriate. In the exercise of discretion, courts have taken the view that considerations of judicial economy favor consolidation. However, the discretion to consolidate is not unfettered. Considerations of convenience and economy must yield to a paramount concern for a fair and impartial trial.
Johnson v. Celotex Corp., 899 F.2d 1281, 1284-85 (2d Cir.) (citations omitted), cert. denied, 498 U.S. 920, 111 S.Ct. 297, 112 L.Ed.2d 250 (1990); see also Flintkote Co. v. Allis-Chalmers Corp., 73 F.R.D. 463, 464 (S.D.N.Y.1977) (“consolidation should not be ordered if it would prejudice defendant”). In Johnson, we noted that a court considering a consolidation must consider:
“[Wjhether the specific risks of prejudice and possible confusion [are] overborne by the risk of inconsistent adjudications of common factual and legal issues, the burden on parties, witnesses, and available judicial resources posed by multiple lawsuits, the length of time required to conclude multiple suits as against a single one, and the relative expense to all concerned of the single-trial, multiple-trial alternatives.”
899 F.2d at 1285 (alteration in original) (quoting Hendrix v. Raybestos-Manhattan, Inc., 776 F.2d 1492, 1495 (11th Cir.1985) (citation omitted)).
The benefits of efficiency can never be purchased at the cost of fairness. As we recently stated:
[W]e are mindful of the dangers of a streamlined trial process in which testimony must be curtailed and jurors must assimilate vast amounts, of information. The systemic urge to aggregate litigation must not be allowed to trump our dedication to individual justice, and we must take care that each individual plaintiffs — and defendant’s — cause not be lost in the shadow of a towering mass litigation.
In re Brooklyn Navy Yard Asbestos Litig., 971 F.2d 831, 853 (2d Cir.1992); see also Arnold v. Eastern Air Lines, Inc., 712 F.2d 899, 906 (4th Cir.1983) (en banc) (“considerations of convenience may not prevail where the inevitable consequence to another party is harmful and serious prejudice”), cert. denied, 464 U.S. 1040, 104 S.Ct. 703, 79 L.Ed.2d 168 (1984); Baker v. Waterman S.S. Corp., 11 F.R.D. 440, 441 (S.D.N.Y.1951) (“a fair and impartial trial to all litigants” is the foremost concern when considering consolidation); Cain v. Armstrong World Indus., 785 F.Supp. 1448, 1457 (S.D.Ala.1992) (new trial warranted where “[a]s the evidence unfolded ... it became more and more obvious ... that a process had been unleashed that left the jury the impossible task of being able to carefully sort out and distinguish the facts and law of thirteen plaintiffs’ cases that varied greatly in so many critical aspects”).
To strike the appropriate balance as to consolidation vel non, “[c]ourts in the Southern and Eastern Districts of New York have used [a standard set of] criteria ... as a guideline in determining whether to consolidate asbestos exposure cases.” Johnson, 899 F.2d at 1286. These criteria include: “‘(1) common worksite; (2) similar occupation; (3) similar time of exposure; (4) type of disease; *351(5) whether plaintiffs were living or deceased; (6) status of discovery in each case; (7) whether all plaintiffs were represented by the same counsel; and (8) type of cancer alleged...Id. (quoting In re All Asbestos Cases Pending in the United States District Court for the District of Maryland slip op. at 3 (D.Md. Dec. 16,1983) (en banc)). As in Johnson, we again conclude that the test furnishes a useful guideline to evaluate consolidation of asbestos cases.
(1) Worksite
Plaintiffs did not all work at the same worksite. Rather, their only worksite similarity was that each was alleged to have suffered some part of his asbestos exposure at one or more of over 40 power-generating plants throughout New York State. Judge Sifton apparently selected the 48 cases based on his conclusion that, in each case, the plaintiff had suffered “primary” exposure in such powerhouses. The basis for this conclusion is unclear. For example, Lewis’s exposure to asbestos came at thirteen worksites, only two of which were powerhouses. Additionally, the work history which his executrix introduced at trial belies any contention that more than 60% of his asbestos exposure occurred at powerhouses. The testimony of Mr. Feeley, discussed above, presents another good example of a plaintiff who spent only a small percentage of his time at power plants. Significantly, he never testified to spending more than a matter of months at any one of the eight or nine plants where he worked during a twenty-one year period.
The work history evidence of other plaintiffs similarly dispels the notion that they shared primary exposure at the 40-odd power plants. Indeed, the record contains evidence of over 250 worksites. Thus, not only was there no common worksite in this case, but any contention that there was a common type of worksite must be viewed with a skeptical eye.
(2) Similar Occupation
This inquiry is significant because a worker’s exposure to asbestos must depend mainly on his occupation. For example, insulators, who actually applied the asbestos, suffered from direct asbestos exposure, whereas sheet-metal workers, like Mr. Lewis, suffered from asbestos exposure in a bystander capacity. The occupations of the plaintiffs in this case ranged from plumbers to machinists to carpenters to boilermakers to sheet-metal workers.
(3) Times of Exposure
The third factor similarly does not support a finding of commonality. The time frame that the jury was required to consider was enormous: a period involving exposures in intervals that began as early as the 1940’s and ended as late as the 1970’s. While some plaintiffs suffered asbestos exposure over periods of up to 30 years, others had much shorter periods of exposure, undercutting the benefit of efficiency, and increasing the likelihood of prejudice, particularly concerning “state-of-the-art” evidence.
(4) Disease Type
Not all plaintiffs alleged the same type of disease. Rather, of the 48 plaintiffs, 28 suffered from asbestosis, 10 suffered from lung cancer, and 10 from mesothelioma. The significance of this disparity is obvious. When the plaintiffs suffer from the same disease, the economy derived by not rehashing the etiology and pathology of the particular disease will be great, while the concomitant prejudice will be minimal. Here, by contrast, the jury was required to hear testimony about three different diseases. The opportunity for prejudice is particularly troubling where, as here, asbestosis sufferers, who may under certain circumstances expect close to normal life spans, are paired for trial with those suffering from terminal cancers, such as mesothelioma and lung cancer.
(5) The Living & The Dead
Some victims in this case were still living during trial. Others had already died. Lewis died in 1985 at the age of 75. The significance of this factor is evident. Drago, 125 F.R.D. at 65-66 (“[T]he presence of wrongful death claims and personal injury actions in a consolidated trial is somewhat troublesome. ... [T]he dead plaintiffs may present the jury with a powerful demonstration of *352the fate that awaits those claimants who are still living.”).
(6) Discovery Status
Keene does not argue that any of the 48 cases was not ready for trial. We note however, the absence of any express finding of readiness in the district court’s decision rejecting a challenge to consolidation. Query: were the 200 third-party defendants that were impleaded two weeks before the trial ready for the trial?
(7) Counsel
Plaintiffs were represented by five law firms, each of which played an active role throughout the trial.
(8) Cancer
Two different types of cancer were alleged: lung cancer, and mesothelioma, a cancer of the lining of the wall of the chest. Each required distinct testimony regarding its etiology, pathology, and consequences.
In addition to the foregoing eight factors, courts contemplating consolidation must also take into account the number of cases affected. In re New York Asbestos Litig., 145 F.R.D. 644, 653 (S.D.N.Y.1993) (consolidating twelve cases). Here, the maelstrom of facts, figures, and witnesses, with 48 plaintiffs, 25 direct defendants, numerous third-and-fourth party defendants, and evidence regarding culpable non-parties and over 250 worksites throughout the world was likely to lead to jury confusion. Kranz quite properly emphasizes the number of precautions the district court took to assure that each case maintained its identity. See Johnson, 899 F.2d at 1285. We conclude, however, that the sheer breadth of the evidence made these precautions feckless in preventing jury confusion.
Plaintiff contends that even if the consolidation was not warranted, the decision below should nevertheless be affirmed because Keene can show no prejudice arising from it. We disagree. ■ At trial, Keene did not dispute that Lewis was exposed to a wide array of asbestos-containing products; rather, Keene disputed exposure to its products. Also, Keene readily conceded that it was known early on that massive, prolonged, and direct exposure to pure asbestos dust was dangerous; but Keene vehemently disputed that bystander exposure, such as that suffered by Mr. Lewis, was known to be dangerous when Lewis was allegedly exposed to Keene’s products.
We are concerned that the jury’s ability to focus on this distinction may have been compromised in this case. While the evidence regarding Lewis’s exposure to Keene’s products was vague, minimal, and heavily circumstantial when compared to the extensive evidence regarding the products of defendant Owens-Corning Fiberglas, the jury apportioned an equal 9% liability to each defendant. This is hard to explain. We conclude that under the unique circumstances of this case, there is an unacceptably strong chance that the equal apportionment of liability amounted to the jury throwing up its hands in the face of a torrent of evidence. Cf. Arnold, 712 F.2d at 907 (“[wjhile peculiar characteristics of the ... cases might afford an explanation” for an apparently incongruous result, reversal of consolidation was appropriate where “the other possibility [that the prejudice arising from an improper consolidation led to the result] simply could not be eliminated”); Cain, 785 F.Supp. at 1455 (ordering a new trial following the consolidated trial of thirteen asbestos-related cases where “[i]t appeared] that the jury simply lumped ... plaintiffs into two categories and gave plaintiffs in each category the same amount of compensatory damages no matter what their injuries”).
Kranz argues that we cannot reverse on the consolidation issue here without turning our back on In re E. & S. Dists. Asbestos Litig. (Brooklyn Navy Yard), 772 F.Supp. 1380 (E. & S.D.N.Y.1991), aff'd in part, rev’d in part, In re Brooklyn Navy Yard Asbestos Litig., 971 F.2d 831 (2d Cir.1992). There, Judge Weinstein was confronted with a number of cases arising from asbestos exposure to workers at the Brooklyn Navy Yard from World War II through the mid-1970’s. Judge Weinstein first divided the case into three phases: “Phase I consisted of cases in which over 90% of plaintiffs’ exposure to *353asbestos occurred in the Yard; Phase II encompassed cases in which between 50% and 90% of plaintiffs’ exposure took place in the Yard; Phase III covered all remaining cases involving Yard exposure to asbestos.” 772 F.Supp. at 1386. 64 Phase I cases were then tried jointly. Subsequently, 15 cases from Phase II and Phase III were also tried together. The second trial involved only three defendants.
After the two trials, Judge Weinstein rejected a challenge by the defendants to the propriety of the consolidation. He noted our opinion in Johnson, and also referred to the Maryland Factors test:
Several of these criteria are present in the Brooklyn Navy Yard cases. A strong geographic nexus tied these asbestos cases together through plaintiffs’ exposure at one worksite, the Brooklyn Navy Yard. The plaintiffs were represented by a few law firms and sued the same former manufacturers and distributors of asbestos-containing products. Extensive overlap in witnesses, primarily former co-workers attesting to product identification at this worksite and medical and epidemiological experts, saved litigants time and money. The years of exposure spanned the period during which asbestos was utilized at the Navy Yard, beginning in the 1930’s through the early 1970’s.
772 F.Supp. at 1388. The defendants elected not to appeal on the consolidation issue, and accordingly we did not consider its propriety. In re Brooklyn Navy Yard, 971 F.2d at 836 n. 1. We did applaud Judge Weinstein “for his innovative managerial skills,” and indeed, the consolidation in Brooklyn Navy Yard is a chiaroscuro study in contrast with the instant case.
The Brooklyn Navy Yard was owned and operated for the entire relevant time period by one entity, the United States government. Because uniformity is a way of life with the military, the commonality of the 64 Phase I cases cannot be overstated. Presumably, asbestos-containing products were purchased by government contract from the same manufacturers and distributors. Determining the identities of those parties at a particular time was a relatively simple endeavor. Thus, the goal of efficiency was attained. Because the yard was used exclusively to build naval warships, and because such ships were regularly produced pursuant to uniform practices, the overlap in testimony relevant to the cases was obvious. Additionally, workers performing similar occupations on different ships during similar time frames were likely to experience exposure in similar ways. Judge Weinstein recognized that 64 cases where over 90% of asbestos exposure had occurred in a common worksite — subject to the uniformity in procurement and practice typical of the Navy — could be jointly tried without confusing or distracting the jury. The efficiency of the consolidated Brooklyn Navy Yard trial is evident from the fact that it lasted only four months, with an additional four weeks for jury deliberations. This stands in stark contrast with the ten-month trial here — particularly when one remembers that this trial involved 25% fewer cases.
The crucial difference between the Brooklyn Navy Yard case and Keene is that here there simply was no primary worksite. Rather there was an alleged primary type of worksite: powerhouses. Thus, instead of one locus where each plaintiff had worked for most of his career, consolidation was premised on the fact that each plaintiff had spent some part of his career at one or more powerhouses. There was no showing that the powerhouses that served as the focal point for consolidation provided anything like the uniformity at the Brooklyn Navy Yard. There was no finding of common ownership, common suppliers or common practices. It should also be noted that at least one of the powerhouses, Indian Point, is a nuclear facility, and, surely, the construction of such facilities must differ dramatically from that of conventional power-generating stations.
Kranz contends that even if the “strong geographic nexus” distinguishes the first Brooklyn Navy Yard trial, the second trial— involving plaintiffs who had less than 90% of their exposure at the yard — is not distinguishable on that basis. This conténüon is difficult to assess without knowledge of the facts of the plaintiffs’ exposure in the second trial. We note that the second Brooklyn Navy Yard trial had only 12 plaintiffs and *354three defendants. Here, by contrast, there were 48 plaintiffs, 25 direct defendants, and hundreds of third-party defendants.
The distinctions between Brooklyn Navy Yard and the instant case lie at the very heart of our holding today. While district courts need not perform any specific rituals or recite any incantations before ordering cases to be consolidated, they must in every instance consider whether the “actions in-volv[e] a common question of law or fact.” Only then can all be assured that innovative and creative efforts to provide compensation to deserving plaintiffs do not violate Federal Rule of Civil Procedure 42(a), which is designed to achieve efficiency without compromising a litigant’s right under the Seventh Amendment to a jury trial.
We do not wish to be understood as condemning all consolidations of asbestos cases. Our holding today is narrow and amounts to little..more than a caution that it is possible to go too far in the interests of expediency and to sacrifice basic fairness in the process. In ordering consolidation we repeat the counsel of Talleyrand, “Pas trop de zéle” — not too much zeal.
Accordingly, the judgment of the district court is reversed, and the matter remanded for a néw trial.