Opinion of the Court
SULLIVAN, Judge:In 1985 appellant was tried by a special court-martial composed of a military judge alone at the Naval Submarine Base New London, Groton, Connecticut. Contrary to his pleas, he was found guilty of wrongful use of cocaine, in violation of Article 112a, Uniform Code of Military Justice, 10 U.S.C. § 912a. He was sentenced to a bad-conduct discharge, confinement for 3 months, forfeiture of $400.00 pay per month for 3 months, and reduction to pay grade E-l. The convening authority approved the sentence, and the Court of Military Review affirmed (one judge dissenting).
This Court granted review of the following issue:
WHETHER THE MILITARY JUDGE ERRED IN DENYING APPELLANT’S REQUEST FOR AN EXPERT WITNESS ON THE ISSUE OF PROPER URINALYSIS TESTING.
We agree with Judge Gladis’ dissent below and hold that the military judge erred in denying this particular request for Dr. Pitts. See generally Caldwell v. Mississippi, 472 U.S. 320, 323 n. 1,105 S.Ct. 2633, 2637 n. 1, 86 L.Ed.2d 231 (1985); Ake v. Oklahoma, 470 U.S. 68, 105 S.Ct. 1087, 84 L.Ed.2d 53 (1985).
Judge Gladis outlined the general context giving rise to the granted issue, as follows:
The Government sought to prove the accused’s wrongful use of cocaine by evidence consisting of the laboratory results of a urinalysis and expert testimony of a Government witness explaining them. The accused denied that he had used cocaine. The laboratory results showed that the accused’s urine tested positive *435for the presence of cocaine. The defense proffers of the expected testimony of the requested defense expert indicated that, although he agreed that the laboratory results indicated the presence of cocaine, he questioned the accuracy of the tests because of discrepancies in the testing procedures and would interpret the results as negative. The military judge denied the defense motion to produce the requested defense expert at Government expense, finding that, although the testimony of the defense expert was relevant, it was not necessary and there was an adequate substitute in the expected testimony of the Government expert. The defense expert did not testify.
Unpub. op. at 2.
The defense request for this witness states:
Subj: REQUEST FOR EMPLOYMENT OF EXPERT WITNESS ICO U.S. v. ET3 CYRIL E. VANHORN USN, 144-70-5804
Ref: (a) Article 46, UCMJ (b) R.C.M. 703, MCM (1984)
1. In accordance with references (a) and (b), the Defense respectfully requests that Dr. Lucius Loring Pitts, II, Ph.D., of 2787 Birchwood Drive, Orange Park, Florida 32073, be employed at government expense for purposes of testifying for the Defense in the subject case. Dr. PITTS is currently self-employed as a Toxicology Consultant. He is retired from the United States Navy as a Lieutenant Commander in the Medical Services Corps. His job in 1983 and 1984 was that of Technical Director at the Naval Drug Screening Laboratory in Jacksonville, Florida. The lab at Jacksonville employed the same standard operating procedures which are currently utilized at the lab in Great Lakes, Illinois.
2. Dr. PITTS has reviewed all lab documentation in this case which has been provided to the Defense by the Government. If called by the defense, Dr. PITTS would testify substantially as follows:
a. That Naval Drug Screening Laboratory, Great Lakes, IL did not follow the standard operating procedures for naval drug testing labs by failing to set the sensitivity threshold properly with respect to the gas liquid chromatography/mass spectrometry tests conducted on ET3 VANHORN’s urine sample. More precisely, Dr. PITTS will testify that lab personnel did not inject enough of the sample to achieve the proper cutoff point of 250ng/ml. Instead, the threshold was set at a lower 150ng standard. Consequently, lab personnel ended up using the raw area to determine sensitivity while using the tangent area to determine the amount. Dr. PITTS will testify that, in using this threshold, the accuracy of the tests conducted was reduced, rendering the quantitative measurement of the metabolites in ET3 VANHORN’s sample suspect.
b. That Naval Drug Screening Laboratory, Great Lakes, IL did not follow the standard operating procedures in that lab personnel failed to investigate the cause of noise registering on the 303 channel during the GC/MS tests conducted on ET3 VANHORN’s urine sample. More precisely, Dr. PITTS will testify that normally readings on this particular channel do not exceed 5 whereas in the tests conducted on ET3 VANHORN’s sample the readings exceeded 15. He will further testify that a noisy 303 channel indicates either that an adequate quantity or quality of gas is not being used, or that the mass spectrometry source is dirty. In either case, Dr. PITTS will testify that such error reduces the accuracy of the tests employed and affects the quantitative measurement of metabolites in the sample.
c. That, together with the fact that ET3 VANHORN’s sample was retested as falling below even a 200ng standard, these defects call into question the reliability of the results of the urinalysis conducted on ET3 VANHORN’s sample by Naval Drug Screening Laboratory, Great Lakes, Illinois.
*4364. Dr. PITTS may be made available by the Defense for the purpose of testifying on a motion to suppress/dismiss. Dr. PITTS fee is $500 per day plus expenses. The Defense would estimate that his testimony would last two days.
The military judge denied this request and made, inter alia, the following findings of fact and conclusions of law:
On or after 6 June 1985, a urine specimen purporting to be that of the accused was tested by a civilian laboratory and found it contained less than 200 nanograms per milliliter of cocaine metabolite.
Ninth, the testimony of Dr. Pitts does not disprove the presence of cocaine metabolite from the tested specimen, only the concentration thereof.
Now with respect to the defense request for Dr. Pitts, the court concludes the following:
First, there is no evidence before this court from which to conclude that the procedures of the Standard Operating Procedure were violated. The utilization of the raw area count in paragraph C of Section XVI does not appear to violate the provisions of the Standard Operating Procedure as that is the sensitivity calibration versus the quantitative calibration.
Further, it appears that the standard requirements of D.3.e. relative to the comparisons of the 303 mass ion peak and the 226 mass ion peak have been complied with. The evidence appears to support the conclusion that the procedures of Section XVI of the Standard Operating Procedures manual were adhered to.
Secondly, the testimony of Dr. Pitts, though relevant, does not meet the added requirements set forth in R.C.M. 703(d), that is, of necessity and of absence of an adequate substitute. The testimony of Dr. Pitts would in large part be cumulative with that [of] Dr. Romberg.
Further, Dr. Romberg, a previously qualified and recognized expert in the field of analysis of urine specimens for drugs of abuse is considered an adequate substitute.
Accordingly, the defense Motion for Production of Dr. Pitts is denied.
Defense counsel shortly thereafter moved for a reconsideration of this decision. He said:
Essentially, Your Honor, what we’re basing our reconsideration on is the recent testimony of Dr. Pitts, that given the procedures — the defects, what he feels to be defects in the procedures, that he would not have reported the sample out as being positive. Dr. Romberg’s expected testimony is that he does not feel that either of the things that Dr. Pitts feels are defects are defects, and that he would have reported out the sample, given the situation of facts here, as being positive. Dr. Pitts will testify that he would not have reported it out as positive, if he was the director of the organization at that time. And he also will testify that if there was not enough sample to retest, that he would have reported it out as being negative rather than positive, based upon what he feels to be defects in the testing procedures.
The military judge, finding as a matter of fact that Dr. Pitts would testify that he would report the test as negative, nevertheless denied the motion.
Article 46, UCMJ, 10 U.S.C. 846, states that “[t]he trial counsel, the defense counsel, and the court-martial shall have equal opportunity to obtain witnesses and other evidence in accordance with such regulations as the President may prescribe.” (Emphasis added.) R.C.M. 703, Manual for Courts-Martial, United States, 1984, further states:
(d) Employment of expert witnesses. When the employment at Government expense of an expert is considered necessary by a party, the party shall, in advance of employment of the expert, and with notice to the opposing party, submit a request to the convening authority to *437authorize the employment and to fix the compensation for the expert. The request shall include a complete statement of reasons why employment of the expert is necessary and the estimated cost of employment. A request denied by the convening authority may be renewed before the military judge who shall determine whether the testimony of the expert is relevant and necessary, and, if so, whether the Government has provided or will provide an adequate substitute. If the military judge grants a motion for employment of an expert or finds that the Government is required to provide a substitute, the proceedings shall be abated if the Government fails to comply with the ruling. In the absence of advance authorization, an expert witness may not be paid fees other than those to which entitled under subsection (e)(2)(D) of this rule.
(Emphasis added.) These military law provisions mirror in part the Supreme Court’s holdings that, as a matter of fundamental due process, an indigent defendant must be provided “the basic tools of an adequate defense.” Britt v. North Carolina, 404 U.S. 226, 227, 92 S.Ct. 431, 433, 30 L.Ed.2d 400 (1971). See generally Moore v. Kemp, 809 F.2d 702, 709-12 (11th Cir.), cert. denied, — U.S. -, 107 S.Ct. 2192, 95 L.Ed.2d 847 (1987).
The prosecution based its case against appellant solely on the results of laboratory tests on his urine and the expert testimony of Dr. Romberg, a chemist at the Drug Screening Laboratory in Great Lakes, Illinois. See generally United States v. Murphy, 23 M.J. 310 (C.M.A. 1987). This prosecution witness worked at the laboratory that performed the challenged tests but he did not personally conduct them. It was his opinion (in accordance with standard operating procedure at his laboratory) that the greater-than-300nanogram readings in these reports indicated that appellant’s urine specimen “definitely contained cocaine metabolite.”
The majority of the Court of Military Review held that the military judge did not err in denying the defense request for its scientific expert because his testimony was irrelevant. Mil.R.Evid. 401 and 402, Manual, supra. This somewhat surprising conclusion was premised on the assertion that Dr. Pitts’ testimony was limited to the extent of the presence of the cocaine metabolite in appellant’s urine and not its absence. The suggestion of the court was that the nanogram count and appropriate cut-off number were matters exclusively related to laboratory certification and the degree of concentration of the drug. Accordingly, they had no bearing on a court-martial’s determination whether the drug in any concentration was present in appellant’s urine.
The court below too narrowly characterized the scope of the requested witness’ testimony and its possible use by the defense. His testimony went directly to the weight or reliability, if any, which should be afforded these test results. See also United States v. Harper, 22 M.J. 157, 163 (C.M.A.1986). The Government’s own witness, Dr. Romberg, confirms our conclusion. He testified:
Q: Do you know what that instruction says with respect to cut off level for cocaine?
A: Okay, in our SOP the cut off level for cocaine metabolite is 300 nanograms per milliliter for all tests.
Q: And what is a cut off level, can you define that term?
A: Okay, in order to call a sample positive it has to have more than the cut off level amount of drug or drug metabolite in it. For example if there is only 200 nanograms of cocaine metabolite in a sample, cocaine would be present, but we would call that sample negative, it has to exceed 300.
Q: Could you explain why?
A: Okay, again we artifically set this amount higher, much higher tha[n] what we can measure. We can measure down to 5 nanograms on some of the tests. We did this to make sure there’s no possibility of an error coming from either noise or some contamination or any other possibility.
*438(Emphasis added.) Moreover, Dr. Pitts’ testimony concerning a second test of this urine by a civilian laboratory could also cast some doubt on the accuracy of the Government’s tests. Finally, the testimony of a trained expert that he is reluctant to certify a sample as positive would reasonably be expected to have some bearing on the untrained factfinder in his determination of the same matter. See United States v. Murphy, supra.
The military judge, of course, did not find that the testimony of Dr. Pitts was irrelevant. However, he found that it was not necessary because his “testimony ... would in large part be cumulative with that [of] Dr. Romberg.” He also found that Dr. Romberg was “an adequate substitute” because he was “a previously qualified and recognized expert in the field of analysis of urine specimens for drugs of abuse.” We have no doubt that Dr. Romberg was an expert in this field. However, the fact remains that Dr. Pitts, also an expert, had no connection with the challenged laboratory and had examined its reports which were used by the prosecution. More importantly, he had a contrary opinion concerning reliability of the test procedures used, results reached, and conclusions based thereon. In short, his testimony favored the defense and could not reasonably be considered cumulative of Dr. Romberg or replaceable by his testimony.
The pivotal question at trial was whether the factfinder would rely on the laboratory test results produced by the Government as a basis for finding appellant guilty of the charged drug offense. See generally United States v. Harper, 22 M.J. at 164. This evidence, coupled with the government expert’s interpretation of these results, was the only evidence of guilt presented by the prosecution. See United States v. Ford, 23 M.J. 331, 337 (C.M.A. 1987). To deny the defense a meaningful opportunity to present its evidence, which challenged the Government’s scientific proof, its reliability, and its interpretation, denied appellant a fair trial. See United States v. Murphy, 23 M.J. at 312. See generally Crane v. Kentucky, 476 U.S. 683, 689-91, 106 S.Ct. 2142, 2146-47, 90 L.Ed.2d 636 (1986); Little v. Armontrout, 835 F.2d 1240,1243 (8th Cir.1987) (en banc), cert. denied, — U.S.-, 108 S.Ct. 2857, 101 L.Ed.2d 894 (1988); United States v. Patterson, 724 F.2d 1128, 1129-30 (5th Cir. 1984).
The decision of the United States Navy-Marine Corps Court of Military Review is reversed. The findings of guilty and the sentence are set aside. The record of trial is returned to the Judge Advocate General of the Navy. A rehearing may be ordered.