OBJECTIVE: To assess the association between self reported drug abuse and syphilis and gonorrhoea among pregnant women, Jefferson County, Alabama, United States, 1980-94. STUDY DESIGN: We analysed a prenatal care database and assessed the association of self reported drug use with seropositive syphilis and gonorrhoea using prevalence rates, multiple logistic regression models, and the Pearson correlation coefficient (r) for trends. RESULTS: Overall, 5.5% of the women acknowledged drug abuse, 1.4% had seropositive syphilis, and 4.8% had gonorrhoea. In a multivariate analysis, drug abuse was associated with syphilis (odds ratio 2.9, 95% confidence interval 1.6, 5.3) but not with gonorrhoea. Trends in the annual prevalence of drug abuse closely paralleled trends in the annual prevalence of syphilis, including simultaneous peaks in 1992 (drug abuse, 9.1%; syphilis, 3.2%). There was no such parallel trend between drug abuse and gonorrhoea. Annual prevalence of drug abuse correlated with the prevalence of syphilis (r = 0.89, p = 0.001) more than with the prevalence of gonorrhoea (r = 0.45, p = 0.201). CONCLUSION: Among pregnant women, an increase in drug abuse was closely associated with an epidemic of syphilis, but not of gonorrhoea. Systematically collected prenatal care data can usefully supplement surveillance of diseases and behavioural risk factors associated with them.