This paper empirically examines hypothesis from the accounting, manufacturing, and strategy literatures about volume-based, complexity-based, and efficiency-based drivers of manufacturing overhead costs. Cross-sectional data from a questionnaire of thirty-seven facilities of an electronics company are examined. Subject to caveats regarding the cross-sectional tests, strongest empirical association is found for volume-related cost drivers. There is not consistently strong empirical association for complexity- or efficiency-related drivers. Explanations for the limited association for the complexity- and efficiency-related drivers include proxy problems with the complexity and efficiency concepts and problems in developing uniform measures of variables across a broad cross-section of facilities.