We aimed at testing potential effects of extraocular bright light via the ear canals on human evening melatonin levels, sleepiness and psychomotor vigilance performance. Twenty healthy young men and women (10/10) kept a regular sleep-wake cycle during the 2-week study. The volunteers reported to the laboratory on three evenings, 2 h 15 min before usual bedtime, on average at 21:45 h. They were exposed to three different light conditions, each lasting for 12 min: extraocular bright light via the ear canal, ocular bright light as an active control condition and a control condition (extraocular light therapy device with completely blacked out LEDs). The timing of exposure was on average from 22:48 to 23:00 h. During the 2-h protocol, saliva samples were collected in 15-min intervals for melatonin assays along with subjective sleepiness ratings, and the volunteers performed a 10-min visual psychomotor vigilance task (PVT) prior to and after each light condition. The evening melatonin rise was significantly attenuated after the 12-min ocular bright light exposure while no significant changes were observed after the extraocular bright light and sham light condition. Subjective sleepiness decreased immediately over a short period only after ocular light exposure. No significant differences were observed for mean reaction times and the number of lapses for the PVT between the three light conditions. We conclude that extraocular transcranial light exposure in the late evening does not suppress melatonin, reduce subjective sleepiness or improve performance, and therefore, does not acutely influence the human circadian timing system.