The authors conducted the current study to determine whether personality predisposes some individuals to develop cancer.The current study examined the role of personality variables in 2224 older women recalled for assessment after routine mammography in a breast screening program. Using a semiprospective design, subjects completed self-report measures of defense style, locus of control, emotional expression and control, self-esteem, trait anxiety, and state anxiety and depression while waiting for medical examination. Multivariate analysis of variance was used to control for known risk factor variables and to examine differences between 3 control groups (normal tissue controls, benign/cystic controls not requiring biopsy, and benign biopsy controls) and 298 breast carcinoma subjects.No differences were detected between breast carcinoma subjects and controls based on measures of mature, immature, and neurotic defense style; locus of control of behavior; emotional expression-in, emotional expression-out, and emotional control; self-esteem; anxiety; or depression.The results of the current study found no evidence to support an independent association between these personality measures and the development of breast carcinoma. [See accompanying article on pages 686-97, this issue.]