Recent theories in social psychology suggest that explicitly measured attitudes are particularly valuable for the prediction of deliberate, controlled behaviour. In contrast, implicitly measured attitudes are assumed to be more important for the prediction of less controlled, more impulsive behaviour. Yet, conclusive evidence for the differential predictive validity of both measures is scarce. We hypothesized that limitations of different control resources would lead to functionally equivalent effects. In Study 1, cognitive capacity moderated the predictive validity of both explicit and implicit attitude measures in a choice task. Self-regulatory resources led to similar patterns for eating (Study 2) and drinking behaviour (Study 3). In addition to the predictive validity of implicit and explicit attitude measures, in Study 3 we more closely investigated the relative contributions of explicitly measured attitudes and general restraint standards as two distinct, but complementing constructs that are dependent on control resources.