Polar bears spend the majority of their lives on the sea ice, where they gain access to seals and mates. In western Hudson Bay, the sea ice melts for three to four months in the summer, and polar bears there are forced onto land. These bears live on their fat reserves for the duration of the iceless period, until temperatures get colder in the fall and freeze up begins. The aggregation of polar bears near Churchill, Manitoba during the ice free period has led to a thriving tourist industry, with a large influx of tourists visiting Churchill in the fall in a six to eight week period, yet little is known about the impacts of this industry on the biology of the bears. This study investigated the effect of tourist vehicles and human presence on the behavior of polar bears over the fall of 2003 and 2004. Overall time budgets were estimated for bears, and the behavior of males and females was compared. Females spent significantly less time lying and more time in locomotion than males. Time budgets were also estimated for bears in the presence and absence of tourist vehicles. Bears spent less time lying and more time in a sit/stand position in the presence of vehicles. Air temperature had no significant effect on the time budgets of polar bears. Tundra vehicle approaches were manipulated to determine effects on polar bear behavior, and to investigate any variables that significantly affected response, including habituation. A response was defined as any sudden whole body movement or change in position or behavior at the time of approach. A total of 25% of all bears responded to the experimental vehicle approach. For bears that responded to approach, the average distance at response was 43 m. The average speed of the vehicle was 0.66 ± 0.02 m/s (range 0.23 to 1.15 m/s). Approach variables that significantly influenced the likelihood of response of a bear to an approaching vehicle included angle of approach and vehicle speed. Direct approaches, in which the bear was in the path of the moving vehicle, had a higher probability of eliciting a response than indirect approaches, in which the vehicle stayed to one side of the bear at all times. Higher speeds of the vehicle increased the probability of a response by a bear. Behaviors of the bear that significantly predicted a response were shifting of the body and smacking of the lips. A playback study was conducted to determine the effects of human induced sound on polar bears. There was no significant effect of human sound on polar bears. Results presented here provide the first experimental evidence of variables in the tourist industry that affect polar bear behavior, and the first evidence of behavioral cues predicting a response to vehicle approach.