This experiment investigated the effect of a delay in informative feedback in computer-managed instruction upon later retention. Experimental groups of undergraduate college students were given a computer-managed test and received informative feedback (1) immediately, item by item (2 second delay); (2) following the entire test (20 minute delay); or (3) 24 hours later. A control group received no feedback. For items incorrect on the computer-managed test and later correct on the criterion test, the performance of both groups receiving delayed feedback was significantly better than the group receiving immediate feedback. Also for these items there was a significantly greater increase in confidence ratings in the 24 hour delay group than in all other conditions. (Author/WBC)