This article provides an econometric meta-analysis of the disparity of results among a large sample of electricity externality studies. Most importantly, the analysis shows that parts of the disparity can be attributed to methodological differences; the abatement cost and top-down damage cost approaches tend to produce higher external cost estimates, ceteris paribus, than does the bottom-up damage cost approach. There are also systematical differences in magnitudes among fuels; as expected some of the fuels (i.e., coal and oil) have more adverse impacts than do the renewables (i.e., hydro, wind and solar). Furthermore, the studies that have addressed the full fuel cycle tend to produce higher externality estimates. However, the analysis carried out here is not sufficient to explain all of the variability in externality estimates. Thus, overall the results suggest that the possibility of making general policy decisions based on the Studies carried Out so far may be limited, implying that existing externality studies may have to be improved in order to become more useful for policy makers. (C) 2003 Elsevier Ltd. All rights reserved.