Light elicits non-visual effects on a wide range of biological functions and behavior. These effects are mediated by a melanopsin-based photoreceptor system that is very sensitive to blue light (440–480 nm) relative to the three-cone visual photopic system. The aim of the current study was to assess the time-of-day-dependent effects of two different wavelength monochromatic lights at 458 nm and 550 nm on human cognitive function. We conducted an experiment in the daytime and nighttime on different days. Twelve subjects were selected, none of whom was either morning-type or evening-type, as assessed by a translated version of the morningness/eveningness questionnaire. The cognitive function was measured by event-related potential (ERP) using an oddball task, and arousal level was measured by the Alpha Attenuation Test (AAT). We found that 458 nm light exposure caused a significantly larger P300 amplitude than occurred with 550 nm light. There was a significant interaction among wavelength, time of day, and electrode site. Exposure to 458 nm light induced a larger P300 amplitude at nighttime than in the daytime at the Fz electrode site. The Alpha Attenuation Coefficient (AAC) at nighttime was higher than in the daytime. Our results suggest that short wavelength monochromatic light can affect the circadian rhythms of cognitive functions, and indicate that these effects are mediated by a melanopsin-based photoreceptor system. This study has extended previous findings in terms of time of day, and higher cognitive function by using an endogenous ERP component, P300.