Climate change models predict an increase in the amount and intensity of winter rainfall in the UK. A series of experiments using both container grown plants and field trials was undertaken, to investigate the effect of winter waterlogging on the growth, development and yield of current varieties of winter wheat. The interaction of waterlogging with other factors such as summer drought, seed rate, and sowing date was also investigated. In the field, waterlogging decreased plant population, and uptake of nitrogen over winter from 50 kg/ha to 25 kg/ha. Plants formed new porous nodal roots, which were able to penetrate below the water level. All the varieties tested had a good ability to recover from damage by waterlogging. Plant dry weights per area of waterlogged plots were 50% of controls at the end of treatment in March, 75% at anthesis and 85-90% at harvest. In all varieties except Xi-19, recovery was due to increased number of tillers per plant, a higher proportion of which survived to form an ear. Xi-19 increased the number of grains per ear and the weight of individual grains. Waterlogging early in the season (autumn and early winter) killed more plants, but allowed longer for surviving plants to recover and compensate. Spring waterlogging caused the greatest decrease in grain yield. Sowing in September allowed autumn and spring tillering, thus two chances for compensatory growth. November sown plants did not tiller until spring, and if waterlogging occurred at this time yields were especially poor. Waterlogging did not make crops more vulnerable to subsequent drought or lodging, but the incidence of take-all and stem-base disease did appear to increase. Despite the lack of disparity in tolerance, varieties did appear to have different responses to waterlogging. Varieties with a higher growth rate and demand on resources due to weaker winter dormancy (Xi- 19) appeared to suffer more leaf chlorosis than those with a stronger dormancy (Claire). Deben had good ability to recover due to its high tillering rate in spring, and had the largest yield. Hereward showed the smallest decrease in yield, but had the lowest yield under control and waterlogged conditions. It was hoped that it would be possible to find a simple screening method to identify tolerant varieties at an early stage. Unfortunately plant appearance during waterlogging was not predictive of final yield. Research into stress tolerance has tended to concentrate on identifying `stress genes' but in this case it appears that multi-gene traits are more important in determining crop yield.