Urine samples were tested for cocaine, opiates, methadone, benzodiazepines, and THC. Nine percent of the samples from the clinic were positive, as were 15 percent from the labor deck. The incidence of cocaine-positive and opiate-positive samples was significantly higher among the labor deck admissions. The results suggest that positive identification of substance use is more likely to occur on the labor deck than in the Women's Clinic. This finding confirms the greater volume of referrals for intervention from the labor deck than from the Women's Clinic. The finding is important for program planning, in that previous efforts had focused on both sites equally in hopes of achieving early intervention in pregnancy in the Women's Clinic. There is more than one possible explanation for the relative lack of positive urine samples in the clinic setting. First, patients arrive at this clinic on a semi-scheduled basis. Because of their fear of being identified as a substance abuser and receiving a referral for social service intervention, they are unlikely to arrive at the clinic with "drugs aboard" because they may have temporarily abstained from abusing substances before their appointments. In contrast, the arrival on the labor deck is an unpredictable event that may even be precipitated by drug use. Second, the population that chooses to come to the clinic for prenatal care may be different from the general population of women who deliver on a labor deck; the women in the clinic are motivated enough to have prenatal care and may be less likely to be using drugs. Third, it is possible that the stress of pregnancy in some women may accelerate the use of drugs in response to the stress and thereby cause an increased frequency of positive urine samples at the time of delivery. 1 table and 2 references