Abstract Growth and yield of Populus were examined in a 5-year test of five spacing-harvest regimes in western Washington. Two hybrid clones (D-O1 and H-11) were planted at two woodgrass spacings (0.18 and 0.3 m) and three wider spacings (0.5, 1.0 and 2.0 m). All treatments were replicated three times in a randomized block design; all plots were fertilized, irrigated, and weeded uniformly. Mean annual harvest yields in the woodgrass treatments did not differ significantly between the two clones (D-O1 and H-11) or two spacings (0.18 and 0.3 m), averaging 6.4 to 7.0 Mg ha −1 (tonnes ha −1 ) over the 5-year period. The highest yield in woodgrass treatments was produced in the second year (first year of coppice), and it declined thereafter. Cumulative growth in the wider spacings was significantly greater than in the two woodgrass spacings. Per hectare yields of clone H-11 in the wider spacings at age 5 were two to three times greater than cumulative yields obtained with annually cut woodgrass, averaging 15.7 to 18.8 Mg ha −1 yr −1 . Fifth-year increments in all wider spacings (0.5, 1.0 and 2.0 m) of clone H-11 exceeded 30 Mg ha −1 . Yields of clone D-01 in the wider spacings ranged from 8.1 to 10.9 Mg ha −1 yr −1 , and thus were also substantially greater than cumulative yields of woodgrass. We conclude that annually harvested woodgrass shows little promise as a viable system for growing Populus for biomass. On the other hand, yields in the wider spacings with a longer harvest cycle were substantially higher than previously expected, especially for clone H-11. Thus, possibilities for application of the wide-spaced regimes and longer cutting cycles appear promising.