The growth of Sitka spruce was examined on a range of soil types: a flushed peaty gley, a podzolised brown earth, a surface water gley, a stagnant waterlogged peaty gley and a non-flushed peaty gley. There was a large difference in productivity between sites, local yield class (LYC) decreasing in the order; 26, 22, 20, 18 and 14 m3 ha-1 yr-1 respectively. The soil water regime was characterised at each site and measurements were made of tree roots, soil oxygen status and site nutrition. Conditions remained well oxygenated in the top 30 cm (Ah horizon) at the flushed peaty gley, despite saturation below 20 to 25 cm during autumn and winter. The lateral flow of oxygenated water through this horizon prevented the development of anaerobiosis and enabled roots to survive down to 40 cm. The freely draining podzolised brown earth dried out very quickly, tensiometers remaining failed for two months. Rooting occurred down to induration at 55 cm. The surface water gley had a freely draining topsoil which remained well aerated throughout the year. Rooting was extensive in the Ap horizon (3 cm to 28 cm depth) and occurred down to 80 cm. There was a constantly high, de-oxygenated water-table in the waterlogged peaty gley, which only dropped below 30 cm depth for 46 days during the summer. Rooting was restricted to a depth of 12 cm. The peaty gley had a de-oxygenated water-table which remained around 15 cm depth during autumn and winter. Roots that grew down the profile during the summer as the soil dried out to around -1 bar, were killed by subsequent waterlogging. The differences in LYC between sites are explained in terms of soil water regime. A significant reduction in yield occurred on all but the flushed peaty gley site as a result of an initial period of growth check. Drought stress had little effect on site productivity at the podzolised brown earth. Low yields on the peaty gley are attributed to the diversion of photosynthate into the replacement of roots killed by soil anaerobiosis.