OBJECTIVES: The Ontario Drug Policy Research Network (ODPRN) published recommendations to restrict reimbursement and coverage criteria of testosterone replacement therapy (TRT) in Ontario. The ODPRN’s budget impact analysis (BIA) evaluated: no reimbursement change (option A), restricting coverage of all forms of TRT (option B), restricting oral and topical forms only (option C), or restricting topical forms only (option D). The analysis assumed exponential growth of TRT expenditures and inappropriate use in 7%-46% of patients. The analysis was limited to drug prescription costs and excluded other costs affected by the recommendations, resulting in forecasted savings ranging from $7-$16 million over a 3-year period. JSS Medical Research performed the BIA from the ministry of health perspective. METHODS: Our hybrid epidemiological and claims-based BIA included costs of TRTs and key cost drivers of physician visits, administering injectable TRTs, testosterone level testing and Exceptional Access Program evaluation and processing. Ontario prescription drug expenditures based on claims data, as well as published literature and expert opinion were utilized. We evaluated the impact of the ODPRN scenarios with and without inappropriate TRT use over a 3-year period. RESULTS: Based on the JSS assumption of all patients qualifying for TRT and taking into consideration key cost drivers, option B would cost $1.01 million; option C $766,000, and option D $252,000. Using ODPRN assumptions of inappropriate use, JSS forecasted savings of $373,000-$13.4 million as opposed to savings of $7-$16 million forecasted by the ODPRN. CONCLUSIONS: ODPRN savings exclude key cost drivers and assume a greater magnitude of inappropriate use. The burden of the policy change could completely offset savings and generate costs of up to $1 million to the healthcare budget. Healthcare policy recommendations based on drug costs alone underestimate the true cost, shifting and in this case creating additional costs to other areas of the healthcare system.