The optimal duration of cytostatic treatment for metastatic breast cancer is still a matter of debate. Possible gain in the duration of remission has to be weighed against the side-effects of treatment. Our aim was to define the optimal duration of cyclophosphamide, methotrexate, 5-fluorouracil (CMF) treatment by studying the time to treatment failure, overall survival and using a Q-TWiST analysis. The treating physician's opinion was asked. The European Organization for Research and Treatment of Cancer (EORTC) Breast Cancer Group conducted a randomised trial in 204 non-progressing metastatic breast cancer patients after induction chemotherapy (CMF) to stop or continue treatment. Progression-free (PFS) and overall survival (OS) were studied. To gain more insight into the burden of treatment-related side-effects, Q-TWiST was analysed. In addition, we asked for oncologists' preferences as patients are likely to be influenced by their physicians' opinion. Continuation of CMF had a significantly longer time to treatment failure (TTF) 5.2 versus 3.5 months (P=0.011). There was no overall survival (OS) difference 14.0 versus 14.4 months (P=0.77). Mean quality-adjusted survival time was equal to 8.4 months for no further treatment and decreased to 7.9 months for continuation of CMF (95% Confidence Interval (CI) of difference equals 0.5+/-2.5 months). Almost half of the oncologists said they would favour continuous treatment for a 3-month gain in time to progression-a difference which was not found in this study. Based on these data, an interruption of chemotherapy (CMF), if this is the wish of the patient, is justified.