The purpose of this study was to investigate the association among daily stressors, cognitive rumination, and fibromyalgia symptoms using time-series methodology and to determine whether autocorrelation was present in the self-report data. Twelve female fibromyalgia subjects monitored their daily level of stressors, cognitive rumination, and fibromyalgia symptoms for 30-35 days. Time-series regression analyses indicated that there was a positive association between previous-day stressors and fibromyalgia symptoms for one subject and between previous-day cognitive rumination and fibromyalgia symptoms for four subjects. For 7 out of 12 subjects autocorrelation was present, and generalized least-squares methods were used with these subjects. These results indicate that ordinary least-squares methods may often not be appropriate for within-subject designs with self-report data. These results also question the often reported stressor-physical symptom association. This study illustrates a useful methodology and analysis to investigate psychosocial-physical symptom associations.