R.E.B. appeals a judgment declaring him the father of M.J.B. At trial, R.E.B. denied having intercourse with the child’s mother, T.A.T., during the statutory conception period. See sec. 891.395, Stats. The mother testified that R.E.B. was her only sexual partner during that time, and introduced an HLA *159blood test report as evidence of his paternity.1 The report provided raw data obtained from the blood samples of the mother, child, and alleged father, and did not exclude R.E.B. as the father. The report also stated that "the relative chance of paternity, assuming a fifty percent prior chance, is 99.54%. Paternity is extremely likely.” In the report, the testing agency did not explain how it computed the relative chance of paternity, did not indicate that it had any factual basis for the assumption of a "fifty percent prior chance,” and did not explain the meaning of a "fifty percent prior chance.” The court admitted the report without accompanying expert testimony pursuant to sec. 767.48(1), Stats., and the mother offered no other explanation of the report.

In closing arguments to the jury, the trial court barred R.E.B.’s counsel from noting that there was no evidence that the testing agency relied on any of the particular facts of this case in assuming a fifty percent prior chance of paternity. We reverse because we conclude that the trial court improperly limited counsel’s remarks concerning the prior chance assumption.

The underlying problem in this case is the lack of any guidance for the jury in understanding and interpreting the blood test report. "Assuming a fifty percent prior chance,” and the effect of this assumption on the resulting statistical conclusion are not matters that the jury could reasonably be expected to understand without some explanation. The assumption is made, without regard to the facts of the particular case, that there is a fifty percent probabili*160ty that the alleged father had intercourse with the mother and conceived this child as a result. See Everett v. Everett, 150 Cal. App. 3d 1053, 1062 (Cal. Ct. App. 1984). This assumption can also be described as a fifty percent chance that the defendant is the child’s father, and a fifty percent chance that a random man is the child’s father. Id. at 1068 n. 15.

The fifty percent prior chance assumption has no scientific basis, but is employed precisely because nothing is known about whether intercourse actually took place between the parties at a time when conception could have occurred. Since the assumption is not based on empirical facts but, rather, is employed to make the paternity formula work, the probability of paternity results might not be reliable in cases where the occurrence of intercourse and the likelihood of conception at a given time are in dispute.

The percentage calculated as the chance of paternity could vary considerably depending on the traits of the individuals tested, if some other prior chance percentage is used. Id. at 1062. For example, if the mother and the alleged father did not engage in sexual intercourse or if the alleged father is sterile, the prior chance would be zero percent and the resulting probability of paternity would correspondingly be zero percent.2 Because the jury must determine whether intercourse took place between the *161mother and the alleged father at all, and whether it occurred at a time when the child could have been conceived, it is improper to allow the jury to take the probability of paternity result at face value without also considering the underlying premise upon which it is based.

Without the jury considering the fifty percent validity explanation in the report, there was a strong probability that it used the report’s statistical conclusion to find that R.E.B. must have had intercourse with the mother during the conception period. That would not be a proper use of the report, however, because the factfinder has then used the conclusion to prove the arbitrary assumption it is based on, rather than first determining whether the assumption is valid. An obvious assumption underlying the HLA paternity test is that the mother and the alleged father have engaged in sexual intercourse at least once during the period of possible conception. See O’Bannon v. Azar, 435 So. 2d 1144, 1147 (La. Ct. App. 1983).

Before the jury can consider the statistical chance of paternity as evidence of paternity, it must first independently determine whether the mother and the alleged father had intercourse during the conception period. The trial court therefore abused its discretion when it did not allow R.E.B.’s counsel to remind the jury that the testing agency assumed unproven facts in reaching its conclusion.

Trial courts can better handle this situation in the future through appropriate jury instructions. When the HLA report provides a probability of paternity result based upon the validity of an underly*162ing assumption (i.e., in this case a fifty percent chance that the mother and the alleged father had intercourse at a time when the child could have been conceived), the jury must be instructed to determine whether intercourse took place between the mother and the alleged father during the presumptive period. If it so finds, it can then give appropriate weight to the probability of paternity.

The probability of paternity calculation does not constitute evidence that the mother and the alleged father had intercourse during the presumptive period. The instructions should also provide, therefore, that if the evidence does not prove to a reasonable certainty by clear, satisfactory, and convincing evidence that the mother and the alleged father had sexual intercourse at a time when the child could have been conceived, then the jury should find nonpaternity regardless of the HLA probability of paternity results.

By the Court. — Judgment reversed and cause remanded for a new trial.