Cancer patients often have to deal with severe side effects and psychological distress during cancer treatment, which have a substantial impact on their quality of life. Among psychosocial interventions for reducing treatment-related side effects, relaxation and imagery were most investigated in controlled trials. In this study, meta-analytic methods were used to synthesize published, randomized intervention–control studies aiming to improve patients' treatment-related symptoms and emotional adjustment by relaxation training. Mean weighted effect sizes were calculated for 12 categories, treatment-related symptoms (nausea, pain, blood pressure, pulse rate) and emotional adjustment (anxiety, depression, hostility, tension, fatigue, confusion, vigor, overall mood). Significant positive effects were found for the treatment-related symptoms. Relaxation training also proved to have a significant effect on the emotional adjustment variables depression, anxiety and hostility. Additionally, two studies point to a significant effect of relaxation on the reduction of tension and amelioration of the overall mood. Intervention features of the relaxation training, the time the professional spent with the patient overall (intervention intensity) and the schedule of the intervention (offered in conjunction with or independent of medical treatment to the cancer patient) were relevant to the effect of relaxation on anxiety. The interventions offered independently of medical treatment proved to be significantly more effective for the outcome variable anxiety. Relaxation seems to be equally effective for patients undergoing different medical procedures (chemotherapy, radiotherapy, bone marrow transplantation, hyperthermia). According to these results relaxation training should be implemented into clinical routine for cancer patients in acute medical treatment. Copyright © 2001 John Wiley & Sons, Ltd.