Light can influence physiology and performance of humans in two distinct ways. It can acutely change the level of physiological and behavioral parameters, and it can induce a phase shift in the circadian oscillators underlying variations in these levels. Until recently, both effects were thought to require retinal light perception. This view was challenged by Campbell and Murphy, who showed significant phase shifts in core body temperature and melatonin using an extraocular stimulus. Their study employed popliteal skin illumination and exclusively considered phase-shifting effects. In this paper, the authors explore both acute effects and phase-shifting effects of ocular as well as extraocular light. Twelve healthy males participated in a within-subject design and received all of three light conditions--(1) dim ocular light/no light to the knee, (2) dim ocular light/bright extraocular light to the knee, and (3) bright ocular light/no light to the knee--on separate nights in random order. The protocol consisted of an adaptation night followed by a 26-h period of sustained wakefulness, during which a 4-h light pulse was presented at a time when maximal phase delays were expected. The authors found neither immediate nor phase-shifting effects of extraocular light exposure on melatonin, core body temperature (CBT), or sleepiness. Ocular bright-light exposure reduced the nocturnal circadian drop in CBT, suppressed melatonin, and reduced sleepiness significantly. In addition, the 4-h ocular light pulse delayed the CBT rhythm by -55 min compared to the drift of the CBT rhythm in dim light. The melatonin rhythm shifted by -113 min, which differed significantly from the drift in the melatonin rhythm in the dim-light condition (-26 min). The failure to find immediate or phase-shifting effects in response to extraocular light in a within-subjects design in which effects of ocular bright light are confirmed strengthens the doubts raised by other labs of the impact of extraocular light on the human circadian system.