We demonstrated previously that several parameters of a morphine discrimination were significantly different in female versus male rats, using a fixed ratio (FR)-l0 schedule of food reinforcement however, this schedule produced a significant bias in reinforcement frequency between saline and morphine sessions in males but not in females. To determine whether this schedule—drug—sex interaction caused the sex difference in discriminability of morphine, female and male rats were trained to discriminate 3.0mg/kg morphine from saline using a variable interval (VI) 15-s/VI 15-s(Phase 1), a VI 7.5-s/Vl 15-s(Phase II), and a VI 15-s/Vl 15-s(Phase 111) schedule of food reinforcement on morphine/saline levers, respectively. After a minimum of 40 training sessions in each phase, mean reinforcement rates in morphine sessions were highest and the ED50values for morphine discrimination were lowest, in Phase II. Thus, as predicted, the morphine dose-effect curves shifted to the left from Phase 1 to Phase II, and back to the right from Phase n to m, presumably due to the bias in reinforcement rate between saline and morphine sessions that was induced by manipulating the VI schedule on the morphine lever. However, there were no sex differences in the morphine versus saline reinforcement rate or in discrimination ED50in any phase, suggesting that the sex difference observed in our initial study was probably due to the bias in reinforcement frequency (towards the saline condition) that occurredonly in malesunder the FR-10 schedule. This study demonstrates the importance of considering group differences in schedule—drug interactions when comparing discriminative stimulus properties of drugs between groups. Behav Pharmacol 1998; 9:357–362 1998 Lippincott-Raven Publishers.