Objective: A recent randomised clinical trial demonstrated that specialist health professionals (SHP) and volunteer peer advisors in diabetes (PAD) are equally effective in delivering training programs for self-management of diabetes, based on a validated five-item knowledge questionnaire. The objective of this study was to compare the costs of training delivered by SHP and by PAD. Methods: Cost of SHP (specialist nurse) in the NHS setting, including their education/training costs, was obtained from published sources. Cost of training of PAD was based on actual resource utilization during their training: 33 sessions (including instruction and mentoring) of 90 min each were needed at cost of time of a SHP trainer per group of 15. Intervention time for groups of 15 was 6 training session of 90 min each, based on the clinical trial. Patient concordance was included in the analysis. All costs were for 2007. As many volunteers discontinue their commitment, assumptions as to their prospective engagement were tested in scenario analysis. Effects of the voluntary involvement of PAD on their own health and cost outcomes, and cost consequences of the effects of training were not considered. Results: The cost per patient completing the training delivered by SHP was £45.19. The cost for PAD depended on the number of courses delivered after they have been trained: for one course scenario, the cost was £21.52, for 3 courses £7.17, and for five courses £4.30. The respective cost savings per 1000 patients completing the training were £23,673, £38,021, and £40,891. Engaging PAD instead of SHP to train 1000 patients would allow the NHS to provide training on self-management in diabetes to additional 1100-9500 patients. Conclusion: Engaging volunteer peer advisors in the training on self-management of diabetes is a highly cost-saving intervention with potentially considerable implications for public health in the United Kingdom.