Florida is one of several states that have sought to protect newborns by requiring that mothers known to have used alcohol or illicit drugs during pregnancy be reported to health authorities. To estimate the prevalence of substance abuse by pregnant women, we collected urine samples from all pregnant women who enrolled for prenatal care at any of the five public health clinics in Pinellas County, Florida (n = 380), or at any of 12 private obstetrical offices in the county (n = 335); each center was studied for a one-month period during the first half of 1989. Toxicologic screening for alcohol, opiates, cocaine and its metabolites, and cannabinoids was performed blindly with the use of an enzyme-multiplied immunoassay technique; all positive results were confirmed. Among the 715 pregnant women we screened, the overall prevalence of a positive result on the toxicologic tests of urine was 14.8 percent; there was little difference in prevalence between the women seen at the public clinics (16.3 percent) and those seen at the private offices (13.1 percent). The frequency of a positive result was also similar among white women (15.4 percent) and black women (14.1 percent). Black women more frequently had evidence of cocaine use (7.5 percent vs. 1.8 percent for white women), whereas white women more frequently had evidence of the use of cannabinoids (14.4 percent vs. 6.0 percent for black women). During the six-month period in which we collected the urine samples, 133 women in Pinellas County were reported to health authorities after delivery for substance abuse during pregnancy. Despite the similar rates of substance abuse among black and white women in our study, black women were reported at approximately 10 times the rate for white women (P less than 0.0001), and poor women were more likely than others to be reported. We conclude that the use of illicit drugs is common among pregnant women regardless of race and socio-economic status. If legally mandated reporting is to be free of racial or economic bias, it must be based on objective medical criteria.