Most elderly lack a synchronized circadian rhythm and often cannot benefit from light therapy. Non-visual effects of light are mediated through intrinsic photosensitive retinal ganglion cells (ipRGCs). With chromatic pupillometry, the functionality of ipRGCs and their functional circuit in the brain can be tested.Keywords: circadian; dementia; intrinsic photosensitive retinal ganglion cells (ipRGCs); pupil; reflex Demented elderly have high rates of disrupted and restless sleep behavior, promoting decay in cognitive performance (Spira, Chen-Edinboro, Wu, & Yaffe, 2014). The application of light therapy for stabilizing circadian rhythm and thereby improving sleep quality has often shown no effect with such individuals, even while their caregivers have reported improved sleep behavior from the same regimen (Friedman et al., 2012). Light is one of the strongest zeitgebers synchronizing circadian rhythm, because of its input to intrinsically photosensitive retinal ganglion cells (ipRGCs) in the eye and the information thus given to the suprachiasmatic nucleus (SCN; Berson, Dunn, & Takao, 2002). One assumption about the failure of this mechanism is that ipRGCs are either insufficiently triggered or damaged, or the subsequent neural circuits impaired. Thus, the SCN, our inner master clock that controls the day-night cycle, fails to receive adequate information from ipRGCs for synchronization. Initial attempts to monitor the function of ipRGCs using chromatic pupillometry showed that specific light stimuli can help distinguish between cone functions with and without ipRGC input by looking at the post-illumination pupillary light reflex (PIPLR), since ipRGCs are related to the PIPLR (Chen, Badea, & Hattar, 2011;Park et al., 2011).To evaluate whether one cause of bad sleep behavior in obviously demented elderly might be a dysfunction in ipRGCs, preclinical testing was performed at two homes for the elderly near Vienna, Austria, first to test its acceptance and feasibility with this population, and then to evaluate individual PIPLRs and optimize the screening method. A total of 31 participants aged 60-99 years (M = 84.2, SD = 8.8) were tested, consisting of 23 females (60-99 years, M = 85.0, SD = 8.4) and eight males (65-95 years, M = 81.9, SD = 9.4). Prior informed consent was acquired from legal representatives, and the procedure was approved by the ethics committee of Munich University of Applied Sciences. In most cases, physicians had diagnosed some form of dementia from observations of behavior and narrations of relatives. No tests for dementia, such as the Mini-Mental State Examination, had been conducted. The nursing staff reported bad sleep behavior, such as disrupted sleep and day napping.For testing the PIPLR, a mobile measurement device was used to stimulate an ipsilateral eye and measure the consensual reflex on the contralateral pupil (Paulick, Novotny, Bachman, & Plischke, 2013). Testing took place on two consecutive days between 9:00 a.m. and 11:00 a.m. Before the start, participants wore the meas...