DECISION
LEWIS, Judge:Despite his plea of not guilty the appellant was found guilty of wrongful use of cocaine by a military judge, sitting alone. The evidence substantiating the alleged use was the result of a urinalysis conducted by the gas chromatography/mass spectrometry (GC/MS) confirmation test method at a contract laboratory. The test revealed the presence of benzoylecgonine, the specific metabolite of cocaine, at a level of 415 nanograms per milliliter (415 ng/ml) of urine. This was well above the 150 ng/ml level required for a positive confirmation test result. We find that the military judge erred in admitting the test result over defense objection of an inadequately developed chain of custody of the appellant’s urine sample, and we reverse.
The chain of custody information placed before the court was contained in the documentation provided to the prosecution by a certifying official associated with the laboratory. Such documentation is commonly referred to as the “litigation package” and, based on our experience in reviewing courts-martial of this nature, consists of such material relating to a urinalysis as is furnished by laboratory supervisory personnel for use during the court-martial process. In this instance, it appears that something less than a full litigation package was available to the parties at trial. Appellate government counsel argue that that which was available was sufficient to overcome a chain of custody objection. We disagree for reasons set forth herein.
The government’s burden of demonstrating an adequate chain of custody is satisfied by a showing that fungible evidence has been preserved in an unaltered state. United States v. Courts, 9 M.J. 285, 290 (C.M.A.1980); United States v. Nault, 4 M.J. 318, 319 (C.M.A.1978). The court must be satisfied by the government’s proffer that in reasonable probability the relevant evidence, the urine sample in this instance, has not changed in important respects. United States v. Courts, supra; United States v. Frost, 19 M.J. 509 (A.F.C.M.R.1984), aff'd, 22 M.J. 386 (CMA 1986). In ruling on chain of custody issues the military judge has broad discretion, and his ruling will not be disturbed absent a clear abuse of that discretion. United States v. Hudson, 20 M.J. 607, 608 (A.F.C.M.R.1985), pet. denied, 21 M.J. 32 (1985). We examine the chain of custody information in this case with this background in mind.
The evidence presented by the government reflects that a portion of the appellant’s urine sample was screened by the radioimmunoassay (RIA) method at the Air Force Drug Testing Laboratory on 6 November 1986. While an RIA screening does not accurately detect the specific metabolite of cocaine, this screening reflected a presumptive positive indication of the presence of the metabolite in accordance with the existing Departmental cutoff standard. Chain of custody documentation available to the court reflects that an aliquot, or portion, of the urine was shipped to the contract laboratory on 12 November 1986. The following day, 13 November 1986, the urine was received at the contract laboratory, assigned a laboratory accession number, and transferred along with certain other specimens to the temporary storage portion of the laboratory. No further documentation purporting to record chain of custody information is apparent until over four months later, on 20 March 1987. This information reflects that the sample was removed from long-term, as opposed to temporary, storage and prepared for the GC/MS confirmation test. This test was conducted three days later on 23 March 1987. We may infer that a portion of the appellant’s sample was actually tested on that date by reference to two identification numbers, the laboratory accession number and the QC batch number reflecting that grouping of identifiable control samples with which the urine was assigned for purpose of the test.
The government attempted to provide an explanation for the gap of over four months in the chain of custody entries. *825In reply to questions by the trial counsel, the government expert, Doctor John Vasiliades, testified that he had contacted a representative of the laboratory by telephone and had received a satisfactory explanation. The military judge properly sustained a defense hearsay objection to the witness’ stating the substance of the explanation. During the course of his testimony Doctor Vasiliades managed to convey the thought that the gap did not represent what he would characterize as a discrepancy. It was rather a question of missing documentation which was not provided with the litigation package. He summed up his thoughts on the matter, constrained as he was by the military judge’s ruling on hearsay, by saying, “Honestly, if there was a problem, ... you would never get a litigation package.”
If we accept the logic of the witness’ observation, certain basic premises emerge. Laboratory personnel assume a quasi-judicial role in determining whether test results have sufficient reliability to support prosecution in a given case. The need for providing detailed chain of custody information ceases to be a concern in this eventuality. The preparation of a litigation package, presumably containing relevant test data as a minimum, ipso facto reflects regularity in processing of the sample tested. Clearly, the adoption of a broadly based presumption supported by the premises set forth would ease the prosecution burden considerably. However, this is not the direction in which the law is heading.
Appellate government counsel argue that laboratory procedures designed to insure proper custodial controls are customarily accorded a superior recognition of regularity to those of law enforcement agencies, citing United States v. Porter, 12 M.J. 129 (C.M.A.1981). We agree that Porter supports that proposition to some extent. See United States v. Strangstalien, 7 M.J. 225 (C.M.A.1979), on which the Porter dictum is based. Strangstalien emphasizes that “a lack of break in the chain of custody” is a key to the presumption of regularity which a court may accord to a laboratory. 7 M.J., at 229. On the other hand, a four month hiatus between the chain of custody entries reviewable in court, in a situation in which the pertinent evidence has apparently been handled in some fashion during the interim, unquestionably represents a break of major proportion. Porter also noted that the accused in that case was contesting his knowing participation in a sale of marijuana and not, specifically, that the substance seized and subsequently tested in a central laboratory was in fact marijuana. Under those circumstances any error in the chain of custody documentation was not prejudicial. 12 M.J., at 133. Herein, the presentation of a traceable chain of custody assumes a more paramount role.
The primary reason that regularity cannot readily be presumed in this case stems from Doctor Vasiliades’ testimony concerning his understanding of the regular practice for handling specimens by the laboratory in question. According to the witness, a sample is placed in temporary storage until such time as testing has been accomplished and testing results are validated. Then, in the normal course of events, the remainder of the sample would be transferred to long-term storage. This facility occupies a separate room down the hall from the temporary storage area, although both facilities occupy a common controlled area, the laboratory itself. Doctor Vasiliades opined that a sample might be placed into long-term storage as soon as it enters the laboratory if the entire space in temporary storage is being utilized. Based on this testimony, one might engage in some creative speculation as to what occurred with regard to the sample between 13 November 1986 when it was placed in temporary storage and 20 March 1987 when it was removed from long-term storage, ostensibly to be prepared for the GC/MS confirmation testing. It is not possible to draw a reasonable inference from this testimony, however. The described customary practices simply do not equate with the combination of entries on either side of the four month gap in the chain of custody paperwork.
We conclude that the prosecution was not relieved of its burden of establishing a *826continuing and meaningful chain of custody by virtue of the military judge’s ruling which precluded its doing so by eliciting hearsay testimony. The military judge, while ruling correctly on the hearsay issue, abused his discretion in admitting the GC/MS test data while reserving the chain of custody question as a “factual issue.” At no point in the proceeding did he have a sufficient evidentiary foundation to discharge his responsibility as a trier of fact of the chain of custody issue in an informed fashion. We conclude that no credible inference that a urine substance has been tested in its unaltered state can be based on a gap in chain of custody documentation of the magnitude presented in this case.
The appellant, a noncommissioned officer with an outstanding record, had been selected from the 702 career field for a position in the Wing Command Section. The record indicates that he was performing in a commendable manner in that position at the time of the positive test result. In testimony on the merits he strenuously denied ever knowingly ingesting cocaine. The only competent evidence to overcome his denial was the GC/MS confirmation test result. The test result should not have been considered in the circumstances of this case. While we do not for a moment suggest that any accused in a urinalysis case is entitled to any less, Staff Sergeant Carrott was entitled to a good deal more than an unexplained four month absence of chain of custody documentation on his day in court. At this point he is entitled to a new day in court.
A separately asserted error need not be decided in view of our disposition. Nevertheless, it merits brief discussion. In rebuttal to the appellant’s testimony suggesting that he had never used cocaine, Doctor Vasiliades was recalled. Having reviewed data from a negative RIA screen of another urine sample provided by the appellant, he testified that, while not sufficient to reach the Departmental presumptive positive cutoff standard, the data appeared to establish the presence of cocaine metabolites when reviewed from a purely scientific standpoint. Since an RIA screening does not isolate the specific metabolite, benzoylecgonine, as we have noted earlier, we assume Doctor Vasiliades was referring to those metabolites generally associated with a prior ingestion of cocaine. We have recently addressed the admissibility of what purported to be a negative drug test result in a much different context. See United States v. Joyner, 25 M.J. 730 (A.F.C.M.R.1987). We urge a thoughtful review of that opinion by all parties in the event this issue should arise in a rehearing.
The findings of guilty and the sentence are set aside. Inasmuch as the record reflects that an adequate chain of custody might yet be established, a rehearing may be ordered. If a rehearing is deemed impracticable, the convening authority should dismiss the charge and specification.
Senior Judge SESSOMS and Judge BLOMMERS concur.