Ecologically relevant multi-temporal-scale frameworks elucidate whether patterns of animal activity (movement) are distinct at different temporal scales. We studied fishers (Pekania pennanti) fitted with transmitter collars that measured activity at high temporal resolution during winter in Wisconsin. A multi-temporal-scale framework with two scales, day (number of activity bouts per 24 h) and decision (time spent exploring a small area), revealed that the best models of fisher activity were different at different temporal scales. At the day scale, the fishers that we studied increased activity when temperatures were warmer and increased activity more when it was windy and cold than when it was windy and warm. At the decision scale, when wind speeds increased, female fishers reduced activity more than male fishers. The role of temperature on activity was only detected at the day scale, and intersexual variation was only detected at the decision scale. Results at both scales supported the prey synchrony hypothesis for activity, but support was mixed at the day scale. Assessing activity using a multi-temporal-scale analysis framework leads to a better understanding of factors that influence activity and biological fitness.