Abstract Background A small share of patients account for a large proportion of costs to the healthcare system in Denmark as in many Western countries. A telephone-based self-management support, proactive health support (PaHS), was suggested for prevention of hospitalisations for persons at risk of hospital admission. These persons have chronic diseases, unplanned hospitalisations and age ≥ 65 years. However, evidence is limited on whether this type of intervention is cost-effective. Aim The aim of this study was to assess the incremental cost-utility ratio (ICER) of PaHS, compared with standard care. Methods The economic evaluation was nested within a randomised controlled trial, and was based on a health system perspective, with follow-up and time horizon of 12 months. We measured incremental costs per quality-adjusted life years (QALY) gained. Total average costs per patient included PaHS programme costs, and costs in hospitals, primary care and municipalities. We analysed differences by generalised linear models with Gamma distribution for costs and mixed models for QALY. Results We analysed data on 6,139 patients, where 3,041 received PaHS and 3,098 received usual care. We found no difference in healthcare costs, and programme costs were on average €1,762 per patient, providing incremental costs of €2,075. Incremental effects on QALY were 0.007, resulting in an ICER of €296,389 per QALY gained. Conclusion We found no evidence of PaHS being cost-effective in this study, but the results will be used to identify new ways to organise similar interventions and identify patients with the objective to reduce health system costs per patient.