The study examined whether the five subscales of the relatively unexplored measure of GWA were reliably measured by a single underlying construct. It also explored whether the two workplace-based measures were distinct but related constructs, as well as the consistency of their factor structure across genders. Date were obtained from a fall 1996 mail survey of a stratified sample of 4,832 employees on the payroll of an urban midwestern university. The sample, divided into 8 gender/occupational strata, yielded a 52-percent response rate, with 1,336 females and 1,156 males participating. The analyses confirmed that SEQ and GWA were two distinct but strongly related constructs. Further, the factor structure of these two constructs, as well as their independence, held across gender. For both men and women, sexual coercion and physical aggression were found to be unreliable measures of their respective constructs, SEQ and GWA. The low reliability of sexual coercion in this study was consistent with other results reported for this variable. The findings suggest that additional research is required to investigate the "low base rate" generated for the extreme behavior items. Although research on negative workplace experiences has focused on sexual harassment experiences, more generalized abusive experiences that do not explicitly focus on gender-based victimization are probably more prevalent and related to a variety of detrimental mental health outcomes. Both measures add to the understanding of the role of the workplace in adverse health outcomes. 5 tables, 2 figures, and 27 references