Aggregated retention is now a common method of regeneration harvest in forest ecosystems managed for both timber and ecological objectives. If residual forest aggregates are to serve as temporary refugia for species sensitive to disturbance or environmental stress, microclimatic conditions must be sufficiently buffered to allow for their persistence. In 1-ha aggregates at three experimental sites in western Washington, we quantified spatial gradients in microclimate (light, air and soil temperature, and soil moisture), effects of aspect on these gradients, and how microclimate compared to conditions in adjacent harvest areas and larger tracts of undisturbed forest (controls). Light availability and temperature were greatest at the edge, but declined sharply inside the aggregate, with most change occurring within 20 m of the edge. Beyond this distance, light generally declined to levels observed in the controls. Soil temperatures exhibited greater spatial variation and stabilized further from the edge (10 to 30 m), but air temperatures were generally higher than those in controls. Soil moisture exhibited no spatial trends and was comparable among aggregates, harvest areas, and controls. Aspect exerted strong effects on light and temperature, particularly within 15 m of the edge, as did forest structure. Where tree density was low, microclimatic gradients were less steep and aspect-related differences were small. Comparisons with previous studies of ground vegetation indicate that microclimatic effects were consistent, in part, with declines among some groups of vascular and nonvascular plants; however, these declines were restricted to edge environments (5 to 10 m) and were unaffected by aspect. Our results suggest that 1-ha aggregates are sufficiently large to contain areas with light, temperature, and soil moisture that are comparable to those in undisturbed forest and suitable, in the short term, for persistence of forest-dependent species.