This secondary analysis study investigated HIV sexual and drug-use risk in drug-dependent pregnant patients over the first month postrandomization to reinforcement-based treatment (RBT) () or usual care (UC) (). Analysis of primary outcomes had indicated that RBT participants spent significantly longer time in treatment and recovery housing than UC participants. The present study examined the ability of 9 risk markers—age, race, estimated gestational age at treatment entry, lifetime substance abuse treatment episodes, history of prostitution charges, history of serious depression, current heroin injection status, current housing status, and current partner substance use—to predict changes in HIV risks. Sexual risk declined for participant subgroups with prostitution-charge histories and unstable housing. Drug-use risk declined for heroin injectors and nondepressed participants. A relationship was found between number of lifetime drug treatment episodes and sexual and drug-use risk. The role of risk markers in the response of drug-dependent pregnant women to drug treatment require attention.