In nocturnal rodents, millisecond light ("flash") stimuli can induce both a large circadian rhythm phase shift and an associated state change from highly active to quiescence followed by behavioral sleep. Suppression of locomotion ("negative masking") is an easily measured correlate of the state change. The present mouse studies used both flashes and longer light stimuli ("pulses") to distinguish initiation from maintenance effects of light on locomotor suppression and to determine whether the locomotor suppression exhibits temporal integration as is thought to be characteristic of phase shift responses to pulse, but not flash, stimuli. In experiment 1, locomotor suppression increased with irradiance (0.01-100 microW/cm( 2)), in accordance with previous reports. It also increased with stimulus duration (3-3000 sec), but interpretation of this result is complicated by the ability of light to both initiate and maintain locomotor suppression. In experiment 2, an irradiance response curve was determined using a stimulus series of 10 flashes, 2 msec each, with total flash energy varying from 0.0025 to 110.0 J/m(2). This included a test for temporal integration in which the effects of two equal energy series of flashes that differed in the number of flashes per series (10 vs 100), were compared. The 10 flash series more effectively elicited locomotor suppression than the 100 flash series, a result consistent with prior observations involving flash-induced phase shifts. In experiment 3, exposure of mice to an 11-h light stimulus yielded irradiance-dependent locomotor suppression that was maintained for the entire stimulus duration by a 100-microW/cm(2) stimulus. Light has the ability to initiate a time-limited (30-40 min) interval of locomotor suppression (initiation effect) that can be extended by additional light (maintenance effect). Temporal integration resembling that seen in phase-shifting responses to light does not exist for either phase shift or locomotor suppression responses to flashes or for locomotor suppression responses to light pulses. The authors present an alternative interpretation of data thought to demonstrate temporal integration in the regulation of phase shift responses to light pulses.