Surveillance of drug safety in pregnancy often draws on administrative prescription registries. Noncompliance in the use of prescribed medication may be frequent among pregnant women owing to their fear of fetotoxic side effects. To estimate compliance in the use of prescription drugs dispensed during pregnancy, we compared prescription data from the North Jutland Prescription Database with information on drug use provided by pregnant women to the Danish National Birth Cohort (DNBC), which is a health interview survey. We used the North Jutland Prescription Database to identify all prescription drugs dispensed during pregnancy for the 2,041 women who were enrolled in the DNBC in the County of North Jutland, Denmark. Compliance was defined as the probability of reporting drug use in DNBC after purchasing a dispensed prescription drug. The overall compliance to drugs purchased within 120 days before the interview was 43% (95% confidence interval = 40-46). Drugs used for treating chronic diseases, for example, beta-blockers, insulin, thyroid hormones, and diuretic and antiepileptic drugs, were always reported to be used, but compliance was low for drugs used for local or short-term treatment such as antihistamines, antibiotics, antacids, nonsteroid anti-inflammatory drugs, and gynecologic drugs. Thus, for the latter drug groups the prescription database may provide an incomplete identification of exposure. Neither data source is unbiased regarding actual drug intake. Nevertheless, our results indicate that for some drug groups risk assessment studies based on prescription data may produce false negative results as a result of noncompliance.