OBJECTIVES: As part of economic evaluations submitted to NICE, probabilistic and deterministic sensitivity analysis are a requirement, with probabilistic sensitivity analysis being a stated preference in the NICE reference case. The aim of including sensitivity analysis is to identify the key areas of uncertainty, and determine the impact on results. The aim of this analysis was to assess what impact uncertainty in cost-effectiveness models has had on NICE reimbursement decisions and review what type of sensitivity analyses are conducted. METHODS: The five most recent NICE appraisals for breast cancer were selected, the sensitivity analysis results and methods were extracted. Once extracted the results of the sensitivity analysis were compared and contrasted. The sensitivity analysis results were considered in the context of the base case results.  RESULTS: The methodology of sensitivity analysis conducted varied between submissions, whilst all appraisals conducted univariate sensitivity analysis only two reported tornado diagrams. The method of reporting results also varied between appraisals, of the four appraisals that had more than one comparator in the base case, only one appraisal conducted a multi-way cost-effectiveness acceptability analysis.  CONCLUSIONS: There are many factors that impact a NICE committees decision, therefore it is not possible to draw a conclusion on how the uncertainty impacted the decision making process. Of the appraisals assessed, there was a wide range of differences between deterministic and probabilistic ICERs, however it appears that this did not impact the appraisal. The sensitivity analysis reported across NICE submissions lacks consistency in the observed sample, hindering the comparison between submissions.