Climate change has substantially shifted the phenology of organisms. These shifts vary across species and habitats and are shaped by species natural history traits and local environmental conditions, yet the relative importance of these drivers remains unclear. Moreover, climate can have diverse effects on different aspects of phenology, such as the timing and duration of activity, but this complexity is rarely captured by commonly used phenological metrics. We used multidecadal butterfly surveys and climate data from five montane sites spanning an elevational gradient to investigate how climate affects different aspects of annual flight period of 135 butterfly species. Using a hierarchical Bayesian framework, we modeled annual probability of occurrence distributions for species using polynomial models that capture changes in abundance, timing, and length of flight. Spring maximum and minimum temperatures and winter precipitation were the best predictors of inter-annual variation in phenology. High winter precipitation, which usually comes in the form of snow, delayed phenology, while warmer spring maximum temperatures advanced phenology across elevations. Even modest increases in spring minimum (nighttime) temperatures caused strong phenological shifts. Climate effects varied among sites, among species within sites, and even among populations of the same species across sites, with particularly pronounced variation among species within a site. Variation in climate effects were slightly better explained by local climate than by natural history traits. Among natural history traits, voltinism and overwintering stage were particularly influential. Importantly, climate influenced different aspects of the flight period (e.g., timing versus duration) in distinct ways, with both natural history traits and local climate modulating these responses. Our findings highlight the often-overlooked importance of winter precipitation and nighttime temperatures in shaping phenology and demonstrate the value of considering the entire flight period, rather than distinct aspects alone, to improve our understanding and predictions of species response to climate change.