Classical conditioning is impaired in elderly subjects, and an experiment investigated the role of unconditioned response magnitude and autonomic arousal in this aging-related impairment. Two groups of young (n = 20, mean age 23.8 years) and elderly (n = 21, mean age 73.8 years) healthy subjects participated in the experiment. Single-cue classical eyeblink conditioning was in effect for 72 trials, with a tone as the conditioned stimulus (CS) and an airpuff to the eye as the unconditioned stimulus (US). The CS was presented alone in eight additional trials to allow recordings of skin conductance and heart rate responses. Startle eyeblink reflexes were elicited either alone or in the presence of the CS after the end of the conditioning procedure to investigate startle potentiation. More young (95%) than elderly (76%) subjects identified the CS - US contingency in a post-conditioning questionnaire, whereas only 38% of the elderly and 85% of the young subjects acquired the conditioned response (CR). Unconditioned blink reflex magnitudes were greater in the subjects who had acquired the CR. Heart rate deceleration was seen in both young and elderly subjects who had acquired the conditioned response. Startle reflexes and skin conductance responses were related to conditioning in young but not in elderly subjects. Unconditioned response magnitude was the best predictor for the acquisition of the CR. Together with the questionnaire data, this indicates that peripheral factors play a significant role in the impairment of classical eyeblink conditioning in elderly people. Autonomic arousal was not related to acquisition of the CR in elderly subjects.