Previous research has demonstrated that contrast effects in multiple schedules occur because of variation in the reinforcement rate following the target components (anticipatory contrast). The present study assessed whether the contingency for transition to the following schedule was an important variable, on the premise that the discriminative cue for the following schedule should be a conditioned reinforcer that should suppress behavior when presented on a responseindependent basis. Strong contrast effects occurred even when the transition to the following schedule was response dependent, and these effects were comparable to those normally obtained with response-independent transitions. Several studies, using both response-dependent (Williams, 1976a(Williams, , 1979(Williams, , 1981Williams & Wixted, 1986;Wilton & Gay, 1969) and response-independent (Farley , 1980; Williams, 1976b) schedules have demonstrated that a major portion of the contrast effect in a multiple schedule is produced by variation in the reinforcement rate following a target component. Moreover, other investigators , using procedures very different from free-operant multiple schedules, have obtained similar results (Bacotti, 1976;Flaherty & Checke, 1982;Flaherty & Rowan , 1985. A critical issue raised by these findings is the mechanism producing such a contrast effect. That is, is the effect (hereafter referred to as anticipatory contrast) adequately characterized as the "primitive" result of the relative reinforcement rate in the target component versus that which follows, or is it possible to identify some more molecular underlying process?Previous research has attempted to identify the attribute of the following source of reinforcement that produced anticipatory contrast. Williams (1979 , Experiment 2) presented pigeons with two separate target components, one of which was followed by a higher valued schedule and the other by extinction (EXT). When these two different following schedules were cued by different discriminative stimuli, large anticipatory contrast effects were observed; that is, response rates were higher in the target component followed by EXT. But when the same cue was used for the two different following schedules, resulting in high response rates in both the higher valued schedule and EXT, anticipatory contrast quickly disappeared . This finding demonstrates that it is not simply the occurrence or nonoccurrence of reinforcement in the following component that is the critical variable. Instead, anticipatory contrast seems to depend upon the "value" of the stimulus in the following component, where value is assessed by the response rate in the presence of the stimulus. One interpretation of such effects of stimulus value assumes that the stimulus correlated with the following schedule possesses conditioned reinforcement properties. It is then possible to view a multiple schedule as similar to other procedures in which behavior is suppressed by the presentation of response-independent reinforcement (e.g., Rachlin & Baum, ...