Variations in longitudinal and radial permeability were investigated in eight seed origins of Sitka spruce (Alaska (AL), British Columbia (BC), Queen Charlotte Islands (QCI), North Washington (NW), South Washington (SW), North Oregon (NO), South Oregon (SO) and California (CA)) grown at two sites in Britain (Dalby, Eastern England and Rhondda, South Wales). Five trees of each seed origin at each site were sampled at three heights (1, 2 and 3 m above ground level). Permeability of timber grown at the two sites showed an inverse relationship between longitudinal and radial permeability. The general trend showed that, although they were less treated longitudinally, the trees on the Rhondda site were more treated radially as compared with those at Dalby. Results obtained in this study generally show that, of the seed origins, QCI and NW had the greatest permeability in either direction. This is in contrast to NO and SO which had the lowest permeability longitudinally (NO) and radially (SO). Therefore, it is suggested that to optimize the permeability in either flow direction, QCI and NW should be selected for plantation use. The seed origins NO and SO should be avoided in future plantations as they were resistant to treatment with liquid preservative. It appears that there are reasonable explanations for differences in longitudinal permeability and that the differences are linked to wood density which is a selection criterion in current breeding programmes.