The statistical issues in clinical trials where clusters, communities or groups rather than individuals are randomized are often not fully appreciated. In this paper we discuss the design and analysis of trials in which pairs of clusters are randomized in the context of one recent trial, the British Family Heart Study. Both sample size calculations and the analysis strategy need to take account of the between-cluster component of variance. The analysis can be considered as a random effects meta-analysis across cluster pairs, and can usefully be presented as such. Techniques developed in the context of meta-analysis can then be used in the analysis, for example using a profile likelihood method to derive a confidence interval for the overall treatment effect which takes into account the variability in the estimate of the between-cluster variance. The methods presented here are contrasted with previously published methods for cluster randomized trials.