This exploratory investigation examined the value of using Lazarus' (1991; Lazarus & Folkman, 1984) stress model, (i.e., primary appraisal, secondary appraisal, and perceived coping) to identify the antecedents of cognitive and somatic state anxiety for endurance athletes. This study also assessed whether endurance athletes with qualitatively similar levels of cognitive and somatic anxiety demonstrate differential antecedent profiles. Participants were 175 triathletes, 70 distance runners, and 70 cyclists who completed stress-related questionnaires 1-2 days prior to competition and the CSAI-2 approximately one hour before competing. Results revealed that all three components of Lazarus' stress model predicted both cognitive and somatic state anxiety better than did individual model components. Moreover, perceived threat accounted for a greater percentage of variance in cognitive and somatic anxiety than did perceived control or coping resources. Cluster analyses revealed distinct antecedent profiles for high, moderate, low, and "repressed" anxious endurance athletes, suggesting that multiple antecedent profiles may exist for highly anxious athletes in endurance sports.