We present a meta-analysis of long-term energy-system scenario studies. The meta-analysis comprises a qualitative taxonomy of modeling approaches and a quantitative decomposition of scenario results across heterogenous studies. The analysis is exemplified by technology-detailed scenario studies of the Swiss electricity system. In the decomposition approach, we assess the variability across scenario results by a principal component analysis, which provides a low-dimensional approximation of multidimensional data. Additionally, by means of a distance measure, the extremality of a scenario result is evaluated, and a minimal set of representative scenarios is determined with respect to a considered scenario result. The proposed methods contribute to the analysis of commonality of modeling approaches and of multidimensional results across heterogenous scenario studies. (C) 2016 Elsevier Ltd. All rights reserved.