Assessing the accuracy of self-reports (SR) of tobacco, alcohol, and illicit drug use among pregnant and post-partum women is crucial to both their treatment and to the health of their children. The public health and drug treatment policy implications of not obtaining accurate self-reports cannot be overestimated. For example, if African-American women are less likely to report drug use than white women, they are less likely to receive treatment. This is a health care disparity issue. Self-report/other alternate measure (drug tests, for example) agreement rates from pregnant women are highly variable according to ethnicity, drug type, and other factors, and no systematic meta-analytic review has attempted to explain this variability. This project builds upon a previous SBIR which has constructed a meta-analytic database containing 325 studies of self-reports of substance use compared to biological indicators (e.g., urinalysis) or collateral reports (e.g., spousal report), and 2083 individual comparisons. A subset of these studies (98) addressing use among pregnant and/or post-partum women have been acquired and are codable. 54 studies have been coded. The present project will summarize the relationship between SR and alternate measures of substance use among pregnant women, the key goals being to provide average estimates of SR agreement, identify distinct subgroups, and search for moderator variables which might influence the relationship between SR and biological indicators in this unique population. The focus will include careful consideration of factors that might be changed to improve self-report accuracy. ? ?