Weather can potentially shape colony development of bumble bees, insects that actively regulate their nest and body temperatures. In this study, I examined the impact of weather (temperature and precipitation) on success of bumble bee colonies using nest boxes set out in semi-natural forested landscapes in the eastern slopes of Southwest Alberta. At each of my 13 study sites, distributed over an 80 x 200 km study area, I installed 12 tree nest-boxes and 12 ground nest-boxes in the spring for use by nest-searching bumble-bee queens, and collected them in the fall. I relate box use and development of colonies that were established in these boxes to weather measured at stations within 20 km of the sites. Colony success was most frequently explained by 2-way interactions of weather variables, particularly precipitation and mean minimum temperature. Colonies were most sensitive to early season weather. Colony success was best explained by two hypotheses explaining colony success: that precipitation enhances flowers, and that minimum night temperatures that are on average warmer reduce costs of colony thermoregulation. Three hypotheses were rejected: that variable temperatures reduce colony success, that colder minimum temperatures reduce colony success, and that rain reduces colony success by preventing foraging. Weather best explained a qualitative measure of colony success and the fraction of pupal cells that emerged, and was less important in explaining the number and the sizes of brood cells. This study highlights the moderate importance of abiotic influences on an important temperate pollinator.