Seventy-eight female breast cancer patients were assessed for fatigue, depression, overall mood, and circadian rhythm at their second and fourth on-study chemotherapy cycles as part of a larger study examining the efficacy of paroxetine in reducing chemotherapy-induced fatigue. The Multidimensional Assessment of Fatigue (MAF), the Fatigue Symptom Checklist (FSCL), the Center for Epidemiologic Studies-Depression [CES-D) questionnaire, the Hamilton Depression Inventory (HDI), and the Profile of Mood States (POMS) were completed by patients at home 7 days after each treatment to assess symptom severity. Circadian rhythm was assessed over a 72-h period with the Mini-Motionlogger Actigraph (Ambulatory Monitoring, Ardsley, NY), starting 6 days after treatment. Daily patterns of sleep and activity were compared across the 3-day period by autocorrelation analyses to calculate a circadian rhythm score for each patient, with higher scores associated with lower disruption. Comparisons of fatigue, depression, and mood with patient circadian rhythm measures taken after the second cycle indicate that all five paper and pencil measures correlated well with the measure of circadian rhythm (all r(partial) <--0.30, all P<0.05). Changes in the fatigue, depression and mood measures from the second on-study treatment to the fourth were significantly correlated with concurrent changes in circadian rhythm (MAF r=-0.31; P=0.04; FSCL r=-0.30; P=0.04; CES-D r=-0.39; P=0.008; HDI r=-0.34; P=0.03; POMS r=-0.40; P=.007). These findings provide evidence that circadian rhythm disruption is involved in the experience of fatigue and depression in cancer patients.