The claim that appropriate "after-event review (AER)" may increase the relative value of drawing lessons from successes, as compared with failures, was examined in the present study. The study was a laboratory experiment in which the effect of type of AER (failure-focused, success-focused, failure- and success-focused, and no AER review) on performance improvement and causal attributions was tested under conditions of earlier success and earlier failure. In general, 2 results were demonstrated: (a) Drawing lessons from successful experience is feasible, and its effectiveness is contingent upon the type of AER. More specifically, after successful events, the most effective review is that of wrong actions, whereas after failed events, any kind of event review (correct or wrong actions) is effective. (b) AERs elicit more internal (as opposed to external) and specific (as opposed to general) attributions. These 2 classifications moderate the effect of AERs on task performance.