Women's experiences with sexual harassment were analyzed with three types of variables: occupational and workplace sex ratios, organizational policies and procedures for dealing with sexual harassment problems, and women's cultural status. Regression analyses revealed that extent of contact with men was a key predictor of incidence of harassment, number of different types of harrassment, sexual comments, sexual categorical remarks, and sexual materials. Gender predominance was a significant predictor of physical threats and sexual materials. Informational methods were less successful than proactive (...) methods in reducing incidents of sexual harassment. The analyses support two generalizations. The “contact hypothesis” tested and verified by Gutek and her colleagues provides a substantive understanding of our findings on workplace and occupational numerical predominance. Second, organizations that take a variety of steps to address sexual harassment are more apt to be successful in curtailing the problem than those relying mainly on “get out the word” techniques. (shrink)