This paper examines the performance of rural-based community groups in Central Kenya and addresses the methodological issues and challenges faced in doing this. Performance measures included subjective and objective ratings of success, including more objectively verifiable measures at household and group levels, derived from a survey of 87 groups and 442 households within four sites. Empirical evidence regarding explanatory factors for relative performance levels is presented using a special sample of 40 groups involved in tree nursery activities, with both descriptive analysis and regression models. Collective action is desired and practised for many tasks. The incredible number, diversity and dynamic nature of groups make it difficult to standardise and measure achievement.Choice and level of performance measures matters in explaining differences in group achievement. Focusing on groups undertaking similar activities allows deeper analysis of performance drivers. Examining different types of groups engaged tree nurseries found that performance was not linked to any easy-to-measure group characteristic, implying that for this task dissemination need not be targeted towards particular types of groups.