Mature-young 5-mo.-old and aged 25-mo.-old rats (N = 28 of each age group) at the same level of motivation were given forced-correct-response training trials on a 14-unit t-maie by blocking off all wrong turns. During later test trials in the maze, when the 14 culs-de-sac were open, aged rats made significantly fewer errors than mature-young rats. The effects of distribution of test trials were relatively minor, but comparison of distributed practice groups with control groups not given forcedcorrect-response training also indicated a strong differential age effect of training procedure on the reduction of errors during test trials. The deleterious effect of age on learning and problem solving is not immutable.