Order of the Family Court, New York County (Leah Marks, F.C.J.), entered September 21, 1989, denying respondent Samuel T. Skeete, Jr.’s motion to vacate, or in the alternative, to renew/reargue an earlier declaration of paternity, is unanimously reversed, on the law, the facts and the exercise of discretion and the motion granted to the extent of holding the prior order of filiation in abeyance and the matter remanded for further proceedings in accordance with this decision, without costs or disbursements.
In this proceeding, initiated in October 1987, the Family Court heard testimony from the parties as to their continuing sexual relationship which began in 1973, resulting in petitioner’s pregnancy in 1980. Petitioner testified as to her exclusive relationship with respondent. Petitioner also testified as to respondent’s mother’s interest in the child (several visits and occasional birthday gifts), and respondent’s sporadic communication with the child on the telephone over the years.
*627In addition to this testimony, the court considered as evidence the results of a blood genetic marker test conducted by Roche Biomedical Laboratories in Burlington, North Carolina. This laboratory was one of the three laboratories certified by the American Association of Blood Banks with authorization to perform such testing for the courts of the State of New York. The samples were drawn in New York and sent immediately to Roche in North Carolina for analysis, using the human leucocyte antigen (HLA) test.
In the test conducted at Roche, blood samples of petitioner, respondent and the child were analyzed. A number of these genetically linked antigens were identified in each sample. One of these, identified as BW53, was found in each of the three samples. Another, identified as A28, was found only in the samples of petitioner and the child, and a third, identified as A3, was found only in the samples of respondent and the child. As a result, based upon statistical probability tables for respondent’s biological group, Roche concluded that the probability of respondent’s paternity vis-á-vis this child was 98.88%.
Relying on these results and the evidence detailing the history of the relationship between the parties, the Family Court made a finding of paternity. This court affirmed unanimously without opinion (149 AD2d 993).
Thereafter, respondent submitted to another series of blood genetic marker tests, including the HLA test, in the Histocompatability Laboratory at Albany Medical College, another of the three labs recognized by the New York courts. This test consisted solely of an analysis of respondent’s blood. However, while the Roche lab had identified respondent’s HLA blood components as Al, A3, B8 and BW53, the Albany analysis agreed with the first three. However, it identified the last B antigen not as BW53 (which matched both the child’s and, incidentally, the mother’s), but BW35.
The Albany lab had the benefit of the Roche results. Thus, it was able to compare the blood analysis of petitioner’s and the child’s samples with the sample of respondent’s blood analyzed in Albany. The A28 match (between petitioner and the child) is, of course, irrelevant to these proceedings. The significance of the A3 match is not clear, because no explanation is provided for the weight accorded each match. But great emphasis is placed on the BW53 match. The Family Court, in its declaration of paternity, did not devote much attention to the component matchups, but instead referred to the over-all HLA test results as another piece of evidence in reaching the
*628determination. The expert testimony before the court at the hearings did not delve into the mechanics of the test and its separate matchup components. However, now that the BW53 link has been seriously questioned, an issue arises as to whether the A3 matchup alone would have been enough to support reference to the probability tables.
The statute permits "one or more blood genetic marker tests” (Family Ct Act § 418 [a]; § 532 [a] [emphasis added]), and whether this language authorizes different kinds of tests or simply retesting for accuracy, we see no great prejudice in a retest under the circumstances here, especially where such a retest would resolve the obvious discrepancy between the two results. The consequences for the parties, and especially for the child, are too serious to warrant ignoring the possibility of error. Concur—Ross, J. P., Carro, Asch and Kassal, JJ.