Assessing the economic impact of medication side effects is critical for the accurate determination of the cost-effectiveness of pharmaceuticals. It is often necessary to employ modeling techniques that utilize incidence rates that are based on the literature. OBJECTIVE: To present a calculation of the cost of steroid induced side effects in renal transplantation as an example of a literature based model and to discuss the assumptions, techniques, and difficulties inherent in the methodology. METHODS: We estimated the incidence of adverse events associated with steroids using evidence-based literature analyses. This evidence was used to construct a Markov cost model for each side effect. Side effects included in the analysis were hypertension (and its complications), post-transplant diabetes mellitus (PTDM), peripheral bone fractures, osteonecrosis of the hip, and cataracts. RESULTS: The side effect incidences reported in the literature vary, in part due to different definitions and follow-up periods. Furthermore, data from randomized controlled trials can be handicapped by low statistical power to detect side effects and patient crossover or dropouts. Confounding and bias are concerns in case-control and cohort studies, and can complicate the determination of attributable risk from cohort studies. Utilizing the incidence data from the literature, the total 10-year discounted cost for a 50-patient cohort was $266,000. CONCLUSIONS: Extracting side effect incidence data from the literature is a challenging but essential step in modeling side effect costs. The variability in reported results should be ascertained, and assumptions requiring clinical expertise are often required. We have demonstrated the methodology and estimated the financial cost of adverse events associated with steroids in renal transplantation.