Using 52 case-comparison studies as the source of data, markers were divided into 4 categories: consistent risk, inconsistent risk, consistent nonrisk, and risk markers with insufficient data. Based on this classification, it appears that a number of widely held hypotheses about husband to wife violence have little empirical support. Findings are discussed in terms of the methodological and theoretical implications of current research on this form of adult domestic violence. 89 references and 4 tables. (Author abstract modified)