Sleep is a critically important behaviour for all mammals due to its fundamental role within homeostatic/circadian systems and memory consolidation. As a large and vigilant prey species that is highly sensitive to stimuli at night, the horse sleeps less than other mammalian species. For this reason, the domestic environment has the potential to greatly affect the duration and quality of equine sleep. This study aimed to determine the effect of environmental factors on equine sleep stages, and whether this would influence cognitive performance during a spatial memory task. Ten riding school horses (mixed breed/ height/ sex; average age 14.9 +2.4 years) were randomly assigned to two groups (n=5) within a five-week crossover repeated measures design experiment. Each group experienced a combination of one of two light conditions (lights on = Treatment; lights off = Control), and one of two bedding depth treatments (15cm bed = control; 5cm bed = treatment) for six days. Duration of sleep stage behaviours (standing Non-Rapid Eye Movement [NREM]), sternal NREM, sternal Rapid Eye Movement [REM] and lateral REM) were measured continuously using CCTV infrared cameras. For the spatial memory task, latency, number of correct responses, and differences between these parameters during training and testing days were measured. A repeated measures general linear model assessed the effects of treatment conditions on duration of sleep stage, and changes in sleep stage over time (bedding and light set as within-subject factors). Wilcoxon Signed-Rank and paired t-tests determined differences in memory task parameters between treatments. Comparing Treatment Bedding with Control Bedding conditions, horses spent on average significantly less time in lateral REM (0.34 ± 0.12 versus 0.46± 0.13hrs; p= 0.032) and sternal NREM (0.64 ± 0.10 versus 0.80 ± 0.12hrs; p= 0.007), and significantly more time in standing NREM (3.69 ± 0.76 versus 3.17 ± 0.77; p= 0.024). Only sternal REM was significantly affected during the Treatment Light condition compared to control conditions (0.53 ± 0.07 versus 0.67 ± 0.11; p= 0.031). Interactions between day and treatment were apparent for specific sleep stage behaviours indicative of acclimatisation. No significant effects (p>0.05) of Treatment Light or Bedding conditions were detected for performance during the spatial memory test. Overall, horses exposed to sub-optimal conditions tended to display significantly less time in recumbent sleep stages (NREM and REM) and increased time in a standing NREM stage. The impact of reduced sleep on equine cognition requires further study.