Circadian rhythms are ubiquitous among eukaryotic organisms and have evolved several times independently which strongly suggests that the presence of a biological clock is adaptive. A prevailing, though relatively untested, hypothesis is that having an internal clock allows organisms to anticipate daily changes in the environment and physiologically and/or behaviorally prepare for them. Locomotor activity is commonly interpreted as an indicator of neurological arousal, and, thus, is a useful indicator in studies of circadian rhythm. We have observed that some species of spiders exhibit a tight and predictable band of activity at a particular time of day, while in other species activity is more spread out through the day. Such patterns tend to continue, or are exaggerated, under conditions of constant darkness. We hypothesize that these predators with tight activity bands are ‘circadian specialists’ using their internal clocks to focus their attention in the narrow temporal window when prey are available to them. On the other hand, the species with broader activity bands are employing a ‘circadian generalist’ strategy, and are able to be more opportunistic with temporally unpredictable prey. Some support of this hypothesis comes from the observation that species that forage only at night tend to have tighter bands of locomotor activity than those species which forage continuously day and night. In this study we present a simulation model that supports this idea that a tight distribution of neurological alertness is advantageous when prey is predictable, but an evenly clumped distribution of alertness is favored when prey are unpredictable. We also tested the abilities of four nocturnal spider species to opportunistically respond to an unexpected period of darkness. Consistent with our hypothesis, we found that in the predicted circadian specialist species activity could not be elicited with unexpected darkness at any time of day, while the predicted circadian generalist showed activity in mid and late day. In two species intermediate on the specialist-generalist scale we only elicited activity in late day. This work represents a test of a novel hypothesis regarding the adaptiveness of circadian rhythms.