A pilot study was conducted to test the efficacy of a therapeutic group by telephone conference call for women with breast cancer. Sixty-six women with stage I or stage II breast cancer consented to participate in the study. Participants were randomly assigned to a usual psychosocial care or intervention group, using a permuted block method. Only 2 of 68 patients dropped out of the study, which included 27% African Americans. Assessments at 3 time periods (pretest, immediately after the intervention, and 3 months after the group ended) included evaluation of quality of life (QOL), mood, and immune function. ttests were performed to determine if differences on important variables existed at pretest. The intervention group had worse QOL and mood scores than did the control group on the pretests. A mixed-model repeated-measures procedure controlling for pretest differences was used to analyze data. A significant Group by Time interaction was found for spiritual well-being and mood. These differences were not in the expected direction. The intervention group showed improvement in QOL and mood during the intervention, but showed decompensation following the intervention. Conversely, the control group demonstrated stable or declining scores. This intervention is feasible and practical for women with breast cancer, especially African American participants. The puzzling results suggest several areas for future research, including a better conceptual fit with outcome measures, increasing dosage, and exploration of the value of emotional expression.