The patterns of seasonal variation of radon concentrations were measured in 91 homes in five regions of the UK over a period of two years. The results showed that there was no significant difference between the regions in the pattern or magnitude of seasonal variation in radon concentrations. The arithmetic mean variation was found to be close to that found previously in the UK national survey. Differences in the pattern between the two years of the study were not significant. Two-thirds of homes in the study followed the expected pattern of high radon in the winter and low radon in the summer. Most of the rest showed little seasonal variation, and a few showed a reversed seasonal pattern. The study does not provide any clear evidence for the recorded house characteristics having an effect on the seasonal variation in radon concentrations in UK homes, though the statistical power for determining such effects is limited in this study. The magnitude of the seasonal variation varied widely between homes. Analysis of the individual results from the homes showed that because of the wide variation in the amount of seasonal variation, applying seasonal correction factors to the results of three-month measurements can yield only relatively small improvements in the accuracy of estimates of annual mean concentrations.