AbstractThe human circadian system is exquisitely sensitive to light, through a pathway connecting the melanopsin-expressing intrinsically photosensitive retinal ganglion cells (ipRGCs) to the hypothalamic suprachiasmatic nuclei (SCN). ipRGCs are characterised by a delayed off-time following cessation of light exposure; we exploited this unusual physiologic property and examined how a sequence of flashes of bright light differing in intensity or duration presented in the biological night could delay the human circadian clock in vivo in healthy young participants (n=54). To understand the mechanism underlying circadian photoreception, we probed temporal integration by manipulating flash intensity and duration independently. In a 34-hour in-laboratory between-subjects design, we examined variable-intensity (3, 30, 95, 300, 950, 3000, or 9500 photopic lux; n=28 participants) flashes at fixed duration (2 ms), and variable-duration (10 μs, 100 μs, 1 ms, 10 ms, 100 ms, 1 sec, 10 sec) flashes at fixed intensity (2000 photopic lux; n=31 participants). We measured the phase shift of dim-light melatonin onset on the subsequent evening, and acute melatonin suppression and alertness during the flash sequence. In the variable-intensity study, we find a clear sigmoidal dose-response relationship for flash intensity and the induced circadian phase shift. In the variable-duration study, we find no parametric relationship between flash duration and induced circadian phase shift, indicating a relative insensitivity of the circadian system to flashes varying in duration. As the intermittent periods of darkness in our stimulation paradigm supports the recovery of extrinsic rod-cone signalling into the ipRGCs, our results strongly suggest rod-cone contributions into circadian photoreception.