Male rats aged 6, 19, or 33 months were trained successively in one- and two-way avoidance tasks. The one-way avoidance test consisted of up to 30 trials given in a single session with the conditional stimulus (CS; 14-kHz tone) presented for either 3 s or for 10 s in separate groups. Senescent rats performed poorest, middle-aged rats intermediately, and young adult rats best. Failure of the longer CS to yield better acquisition than the short CS in the senescent group suggested that the age-related deficit probably did not result from slower responding. In subsequent shuttle box training there was no appreciable age difference in achieving the learning criterion. A compound visual-auditory CS was used, and in further evaluation of well-trained rats it was found that the auditory component was much more effective than the visual component in eliciting avoidance. However, this differential effect of the two stimuli was much weaker in the senescent group than in the young adult group. Nonetheless, these same senescent and adult rats readily learned to make avoidance responses using only the auditory CS, demonstrating that this was an effective stimulus for all age groups.