OBJECTIVES: Previous reviews of cost-effectiveness analyses (CEAs) of varicella vaccination have concluded that dynamic transmission models should be used to properly account for indirect effects of vaccination.  This study reviews CEAs of varicella vaccination that used a dynamic model to identify parameters with the greatest impact on the CEA results.   METHODS: A targeted search of MEDLINE was conducted for childhood varicella vaccination CEAs that used a dynamic model. We assessed the structural assumptions and input parameters that had the greatest impact on the cost-effectiveness results and summarized the ranges of input values and primary data source. RESULTS: Fourteen varicella vaccination CEAs using a dynamic transmission model were identified. The results of these studies were most sensitive to 2 structural assumptions: 1) inclusion of zoster and 2) inclusion of indirect costs. For example, whether or not zoster was included changed results from “cost saving” to exceeding country-specific thresholds for cost-effectiveness. These results depended on the assumed magnitude of the impact of varicella vaccination on zoster cases and the time horizon for the CEA. Three input parameter values also substantially impacted the results: 1) vaccine-related costs depending on if 1 or 2 doses of vaccine were evaluated, 2) estimates of productivity loss per case of varicella (0.27 to 8.8 days for caregivers of children, 2.6 to 26.1 days for adults) and 3) estimates of QALY loss per case of natural varicella (0.0027 to 0.004 [uncomplicated case], 0.0038 to 0.017 [complicated case]), per case of breakthrough varicella (20% to 50% of natural case value), and per case of zoster (0.01 to 0.12  [younger/less severe case], 0.201 to 0.52 [older/more severe case]). CONCLUSIONS: Future research should be prioritized for epidemic and economic parameters for which there is large uncertainty and that impact the results and, consequently, decisions about varicella vaccination programs.