In the context of an epidemiologic multicentric study about perinatal transmission of HIV, screening was systematically proposed to all pregnant women attending nine maternities clinics of the Paris region (n = 7600, between August 1987 and July 1988) at their first prenatal visit. Among them, 2145 had already been tested and 45 were known to be HIV positive. So, 5660 tests were performed during the first 6 months (period 1), and 17 pregnant women were discovered to be positive. The total cost of screening has been estimated between FF. 720,000 and 775,000, resulting in a mean cost per pregnant woman found to be HIV positive of about FF. 42,000 to 45,000. A similar calculation over the following 5 months (period 2) gave a mean cost of between FF. 165,000 and 178,000. Since the women, before being screened, had to answer a short questionnaire about risk factors; cost and effectiveness of a selective screening strategy could be simulated. The preference of systematic screening to selective screening enabled the discovery of two HIV positive cases in each period, the marginal cost, i.e., cost per extra pregnant woman found to be HIV positive was thus FF. 303,320 to 327,540 for period 1, and FF. 572,240 to 619,000 for period 2. Although these figures seem high, an estimation of the cost-effectiveness does not allow us to conclude whether it is in society's interest to devote the funds necessary to move away from selective screening towards systematic screening.(ABSTRACT TRUNCATED AT 250 WORDS)