The idea that merely viewing nature can promote stress recovery has profound societal implications, driving the use of nature imagery in the design of environments where access to real nature is limited. Scientific support for this concept originates from a pioneering study by Ulrich et al. (1991), which demonstrated superior psycho-physiological stress recovery during exposure to nature videos compared to urban videos. Despite its influence, this foundational study has never been systematically replicated. To address this gap, ten research teams across the Netherlands, Belgium, the UK, Sweden, and the USA collaborated on a multisite replication. A final sample of 959 participants (49% women, 51% men, mean age 22 years) were exposed to a 10-minute stressful video followed by viewing a video one of six environments: natural settings (forest or stream), urban pedestrian areas (quiet or busy), or urban traffic areas (quiet or busy). Affective states (ZIPERS; Zuckerman, 1977) were measured at baseline, post-stressor, and post-environmental video, while sympathetic (SRC, PEP) and parasympathetic (RMSSD, RSA) responses were continuously recorded using the VU-AMS device. Results partially confirmed the original findings that psychological recovery was greater for the natural conditions. For physiological measures, only parasympathetic responses showed consistency with the original study in promoting greater initial recovery. Unexpected differences were observed between the two nature videos, potentially due to the loud noise of fast-streaming water in one video. In general, this multisite experiment supports the psychological relevance of nature imagery for stress recovery, while highlighting important nuances in physiological responses.