This experiment queried the frequency with which sexual inactivity in aged male mice is caused by a specific failure to arouse. Level of arousal was assessed in 30-month-old CB6F1 males by measuring each individual's willingness to walk on a treadmill in order to stay close to a receptive female and, in a separate test, by determining the amount of corticosterone secreted in response to a female. Sexual capacity was examined in a third test by challenging each male with a series of receptive females. Absolute time spent near a female on a treadmill, a measure of both endurance and arousal, correlated strongly (0.82) with the frequency with which an old male achieved ejaculation in its test of sexual capacity. Another treadmill measure, the proportion of a male's time spent near a female, correlated less well with ejaculation frequency (0.45) but this measure probably underestimated the importance of arousal. The magnitude of the corticosterone response to a test female also correlated significantly with ejaculation frequency (0.44). When viewed in toto, these results suggest that specific variation in the capacity of aged male mice to experience arousal accounts for 20% to 67% of the variation in their total sexual performance.