Deer-vehicle accidents are an increasing problem in many regions of the world. To elaborate effective mitigation measures, it is necessary to determine environmental factors associated with the occurrence of such accidents. The Lauren-tides Wildlife Reserve in Quebec, Canada, is a prime example of an area having a long-lasting problem of moose-vehicle accidents (MVAs). We tested the effect of the spatial and temporal variables most likely to influence MVAs in this area based on accidents recorded over a 13-year period. Data collected included the date and time of each collision as well as the position of the accident relative to the closest 1-km road marker. We calculated 10 variables to assess moose habitat suitability, moose density, topography and road design for every 1-km road segment. There was a total of 754 MVAs during 1990–2002. The period with the highest number of accidents was the second half of June but accident frequency remained relatively high from mid-May to late August. The risk of accident per vehicle was at least 2–3 times higher at night (when traffic volume was lowest) than during any other time of the day. Also, MVAs were over 42% more frequent on Fridays, when road traffic levels were highest. The probability of an MVA increased when air temperature and atmospheric pressure were high. The MVA rate increased with moose density, in the presence of at least one brackish pool (by 80%) and when a valley traversed the road (by 120%). Implications of our results in the choice of appropriate mitigation measures are discussed. Future work should aim to describe habitat characteristics at the actual collision site.