The studies were all conducted since 1980. They included prevalence rates based on maternal reports, biological measures, and existing data sources. The studies that relied on maternal reports underestimated the problem due to underreporting for social, psychological, and legal reasons. Biological measures produced a wide variety of results due to the time, sensitivity, and specificity of the test. Both maternal and biological measures often could not be generalized to the typical pregnant women, because they examined only certain population segments such as minorities or women at high risk for drug abuse. Moreover, national samples relied on existing sources and did not provide information about the type of drug used. However, an accurate portrayal of the population is essential to the provision of effective prevention and intervention services for pregnant women who use drugs. From such data, non-punitive social policies could be developed both to meet the needs of pregnant drug abusers more effectively than do current policies and to use limited public funds appropriately. Tables and 31 references