The role of anxiety, stress, and social support in predicting negative obstetrical outcomes was examined in a high-risk group of pregnant women. The predictor variables were assessed with separate self-report scales, including The Sarason Life Experience Survey, the Spielberger State/Trait Inventory, and a modified version of the Lazarus and Cohen Social Support Questionnaire. A total of 52 women between their twenty-second and twenty-eighth week of gestation, who were referred to a specialized clinic on the basis of certain medical risks, participated in the study. Data were analyzed in a prospective multiple regression design to investigate whether patients scoring high on anxiety and stress and low on social support were more inclined to experience obstetrical complications. Significant relationships were found between the psychosocial variables and obstetrical outcome. To investigate the possibility that findings were spuriously produced by the diagnosis and referral of patients to the high-risk clinic, psychosocial data collected at the same university clinic on low-risk pregnant women were compared with the data from the high-risk sample. The idea that a high-risk diagnosis leads to higher levels of psychosocial stress or anxiety was not substantiated. The predictive ability of psychosocial variables suggested that these constructs could be included in prenatal screening for better prediction of high-risk pregnancy. (Author/RH)