OBJECTIVES To evaluate the impact of a hypothetical renal health program (RHP) designed to delay initiation of renal replacement therapy (RRT). METHODS We used a Markov model and Monte Carlo technique to simulate two simultaneous 1000 patient cohorts assigned to a renal health program (RHP) or to “standard care” (SC). Population consisted of patients with hypertension and/or diabetes, with initial glomerular filtration rate (GFR) between 60 and 15 ml min (stages 3 and 4 of chronic kidney disease [CKD]). We assumed that RHP had a monthly cost of two to four times that of SC, and that it reduced GFR loss in a proportion that we called “the RHP effect”. The model incorporated 47 variables: 20 epidemiologic variables, 19 cost variables (all at local Colombian rates) and 8 quality of life variables (we used QALY established through experts' consensus). We used a third-party payer perspective, a 5-year time frame and a 3% annual discount rate both for costs and for utilities. RESULTS Assuming a RHP 0.8 effect (equivalent to a 20% GFR loss reduction) the average cost per patient was US$ 17,680 in RHP and US$ 15,560 in SC. Average utility was higher in RHP (3.53 QALY) than in SC (3.45 QALY). From the total of 1000 patients of the RHP group, 101 died compared with 124 in the SC group. Additionally, along the five-year period, 242 subjects from the RHP group required RRT compared with 310 from the SC group. Based on these results, US$ 90,950 would be paid per death averted, US$ 27,200 per RRT averted and the incremental cost-effectiveness ratio would be US$ 21,450 per QALY gained. CONCLUSIONS Five years is a short time. Return to investment would be seen only in patients with advanced CKD. However, a RHP with a modest effect would still de cost-effective.