OBJECTIVES: Analyses of drug cost-effectiveness have traditionally been based on two key components: effectiveness measured by clinical endpoints reported in product label or biomedical literature, and cost obtained from available pricing sources.  While the former is a relatively well-defined value derived from pivotal trials and subsequent studies, the latter has often not been well-described or rigorously assessed, and may be based upon manufacturer-reported pricing information that has been shown to be inaccurate and unreliable. This study examines existing benchmarks against newly available sources of pharmaceutical pricing, describes their interrelationships and assesses their stability over time. METHODS: Using publicly available cost data for wholesale acquisition cost (WAC), average manufacturer price (AMP), national average drug acquisition cost (NADAC), AMP-based federal upper limit (FUL), and a composite of six states’ average acquisition cost (AAC), mean, median, and standard deviation values were calculated for a broad range of pharmaceuticals and trade classes (brand vs. generic) and analyzed temporally for trend and consistency patterns. RESULTS: Available pricing data reflects an extremely high degree of variability and inconsistent relationships, both between equivalent products and from one price type to another.  Generic drug prices demonstrated the greatest irregularity, and although the ratio of NADAC or AMP to WAC for branded drugs showed correlations overall, for given products those relationships could be substantially at variance. CONCLUSIONS: Continuing review and analysis of all available price types is needed to identify a reliable drug pricing benchmark that permits reviewers and clinicians to determine the optimal course of treatment.  Since reported ratios may shift over time, future reports of cost-effectiveness must explicitly identify the cost basis employed and its reliability for the products at issue.