Background: We describe an approach to estimating the cost-effectiveness of an intervention that changes health behaviour. The method captures the lifetime costs and benefits incurred by participants in an ongoing cluster-randomized controlled trial of an intervention that aims to change health behaviour. The existing literature only captures short-term economic and health outcomes. Methods: We develop a state-transition Markov model of how individuals move between different health behaviour states over time. We simulate hypothetical data to describe the costs and health benefits of the intervention, illustrate how the data collected in the ongoing randomized controlled trial can be used and demonstrate how incremental cost-effectiveness ratios are estimated. Results: On the basis of the simulated (i.e. hypothetical) data, we estimate the cost per quality-adjusted life year. The estimate reflects the lifetime health and economic consequences of the intervention. Discussion: The method used for the cost-effectiveness analysis described in this paper is appropriate for investigating whether interventions that change health behaviour in relation to chronic diseases represent good value for money as compared to alternative uses of scarce healthcare resources.