We assessed the impacts of biomass removals, soil compaction and tillage, and vegetation control on early growth of Douglas-fir planted on a highly-productive site in coastal Washington. Harvest treatments included removal of commercial bole (B0), bole only up to 5-cm top diameter (BO5), total tree (TT), and total tree plus all legacy coarse woody debris (TT+). Vegetation control (VC) effects (0 and +) were tested in the BO treatment, while soil compaction (SC) and compaction plus tillage (SC+T) were imposed in the BO+VC treatment. Five years after planting, tree growth was chiefly limited by competing vegetation. Mean tree bole diameter (DBH), total height (TH), and volume index (SVOL = DBH² x TH) in the BO treatment without VC were 34 mm, 308 cm and 4337 cm³, respectively, compared to 44 mm, 356 cm and 8297 cm³ in the BO+VC (P £0.01). Biomass removal and soil compaction effects on tree growth were relatively small. Mean tree SVOL in the BO treatment was greater than in the TT treatment (P <0.05). Mean tree TH was similar in the BO and SC treatments, but increased in the SC+T treatment (P <0.05). Understory vegetation significantly reduced soil volumetric water which was positively related to tree DBH, TH and SVOL. Volumetric soil water content during the low-rainfall mid-summer period appeared to be the main factor limiting tree growth. Supporting this hypothesis, δ¹³C, an index of drought stress, was 0.7‰ lower in latewood of trees in the BO-treatment than of those in the BO-VC treatment (P = 0.02). Although soil solution N concentrations were greater in BO than in TT during years 3 and 4 after planting, foliar N concentration(1.58-1.69%) did not differ among treatments. Treatment effects were related to changes in available soil water rather than to nutrient cycling.