Timing (“phenology”) of flowering within a season can influence reproductive success and produce population-scale mating patterns. Flowering phenology is often partially dependent on environmental cues. Spatial variation in these cues within a population, e.g., due to microclimatic variation, can produce spatial structure in flowering time. Strong spatial structure in flowering time can produce temporal isolation for individuals in different environments, limiting sexual reproductive success or outcrossing, or creating genetic structure. Temporal structure in flowering phenology due to spatial variation in cues has been studied in alpine systems, where microtopography creates variation in the timing of snowmelt. However, environmental heterogeneity and spatial structure in flowering time has not been studied in subalpine meadow systems, despite being a model system for studying flowering phenology and being similarly threatened by climate change. Meadow systems are defined by contrast between open meadow interiors and forest canopy, producing variation in sunlight and soil characteristics. We answered the following questions: (1) does canopy coverage influence the flowering phenology of Thermopsis divaricarpa, a leguminous perennial common to the North American Rocky Mountains, and (2) what are the relative influences of canopy coverage and phenology on seed set in T. divaricarpa.  We monitored flowering phenology over three summers in a large population (>10,000 individuals) within a six-hectare portion of an aspen-conifer meadow complex in Colorado, USA. We performed phenological censuses bi-weekly in 206 square meter plots over three years, recording the number of inflorescences and open flowers for each individual ramet in plots. Canopy coverage for each plot was estimated with a UAV-derived point cloud. In one year, we collected fruit from monitored plants, counting and weighing fruit and seeds. We evaluated relationships among canopy coverage, initiation of flowering, and seed set using spatial autoregressive regression models. In total we observed 490 T. divaricarpa ramets in 103 plots. Across years, canopy coverage delayed flowering, with plants in forest habitat flowering up to two weeks later than plants in open meadow interiors. However, plants in meadow interiors still had high variance in initiation of flowering date, such that there was still temporal overlap in flowering between plants in meadow interior and forest habitats. There was no relationship between initiation of flowering and seed set and only a weak negative relationship between canopy coverage and seed set. These results suggest that subalpine meadows show spatial structure in flowering similar to alpine systems, but it is not strong enough to reduce reproductive success or create strong genetic structure within the population.