Local soil temperature has the potential to affect plant phenology, which is a key indicator of the biological effects of climate change. Many existing analyses, however, ignore local temperatures and focus only on temperature at larger scales measured by weather stations. Ignoring local temperature adds noise to analyses, creating the need for longer time series, and may also bias results. Over four years, we investigated the effects of local soil temperature, sun exposure, and soil depth on flowering phenology for 35 populations of Canada mayflower (Maianthemum canadense) in an 82-hectare mixed deciduous forest in Newton, Massachusetts (USA). Flowering dates varied by 5-7 days among sites. Soil temperatures varied by about 5 °C across sites before and during the flowering season. Among the populations, plants flowered earliest at sites with the warmest local climates-around one day earlier for each 1 °C warmer temperature. Sun exposure and soil depth did not affect flowering times. Differences in temperature and flowering times among sites were consistent over the four years of the study. In most other published phenology studies, spring wildflowers flower 2-5 days earlier for each 1 °C of warming of air temperatures. This study demonstrates that the effects of local temperature on phenology can be investigated over relatively short periods of time and that these effects may bias estimates of phenological responses to temperature that rely solely on temperature data from weather stations.