Aim: Policy-makers typically track the rapidly evolving U.S. residential photovoltaic (PV) market by relying on price data reported by PV installers/integrators to incentive programs. Recent years have witnessed a shift toward third-party-owned (TPO) business models, in which the absence of a cash purchase price obscures data interpretation. Appraisals-often based on estimates of the average fair market value across a diverse fleet of systems-are one way TPO prices are reported. Scope: This study investigates residential PV system price drivers to improve the accuracy, consistency, and relevance of PV price-tracking efforts. Our econometric approach evaluates system price drivers using California Solar Initiative data, controlling for system, installer, and geographic variables. Conclusions: We find that reported prices for confirmed appraised systems are $1.13/W higher than non-appraised systems and do not respond to hypothesized price drivers. For non-appraised systems, we find preliminary evidence of market distortions based on the impact of the incentive level, module cost and household income on reported price. Further, unspecified installer heterogeneity-possibly due to differences in products, cost structure or reporting practices-is a substantial price driver. Using estimates, we develop a price model to approximate non-appraised system prices. (C) 2013 Elsevier Ltd. All rights reserved.