The performance of the NOAA High Resolution Rapid Refresh (HRRR) numerical weather prediction model for resolving low-level winds over a wind energy production site in central California during summer 2019 is evaluated herein. This study intends to catalogue the ability of a high-resolution model to capture boundary layer dynamics relevant to wind energy interests in an area with complex terrain, which has long presented challenges for weather models and wind energy forecasting. Performance is evaluated by comparing HRRR model output to two in-situ wind-profiling Doppler lidars and a surface meteorological station at Lawrence Livermore National Laboratory Site 300. HRRR captured the diurnal cycle of surface layer winds (<80 m above ground level) well, although the horizontal wind speed in this layer was underpredicted during the daytime by an average of approximately 1.5 m s −1 but were well predicted during nighttime hours (mean error < 0.5 m s −1). Above this layer, wind speeds were well predicted during daytime hours (mean error < 0.5 m s −1) overpredicted during nighttime by an average of 2.4 m s −1. A particular challenge for HRRR was the resolution of near-surface (<40 m agl) speedup events, with significant underpredictions of horizontal and vertical wind velocities and turbulent kinetic energy. HRRR model bias relative to observations was found to be minimum during days with synoptic-scale troughs and strong 850 hPa geopotential gradients, while HRRR bias was strongest during days with synoptic ridging and weak 850 hPa geopotential gradients. To translate model performance for resolving winds to energy forecasting, generic wind turbine models were used to estimate power generation from wind speed for characteristic turbines in the nearby Altamont Pass Wind Resource Area. Results show that HRRR-based energy forecasts generally predicted daytime power generation adequately (accuracy <0.4 MW from 09:00 to 14:00 local time for an 18 h forecast), but presented strong overpredictions (>1 MW) over the rest of the diurnal cycle for all forecast hours. These results demonstrate conditions under which HRRR performs well in complex terrain, as well as highlighting potential areas requiring further investigation to support usage of a high-resolution model for wind energy forecasts.