Aims and background The results of several randomized trials and meta-analyses have been reported on adjuvant treatment for early breast cancer and treatment guidelines have been defined accordingly, but detailed data are lacking on the appropriateness of treatment prescription in clinical practice. Methods We performed a prospective, observational, multicenter study to monitor the prescription, delivery and effectiveness of radiotherapy following conservative surgery for early breast cancer; 1610 patients treated with postoperative radiation to the breast in 1997 were entered by 12 centers in Lombardy, Italy. Here we report the results of a secondary analysis focused on the prescription of medical adjuvant treatment (1547 eligible patients). Results Chemotherapy only was prescribed to 526 patients (33%), hormonal therapy only to 539 (33%), and both treatments to 85 patients (5%); 460 women (29%) received no medical adjuvant treatment. We compared the collected data with guidelines defined in 1995 by the St Gallen Consensus Conference. Undertreatment was most frequent in node-negative patients at intermediate/high risk, no treatment (instead of tamoxifen or chemotherapy) being prescribed in 21–45% of cases. Node-negative patients at low risk, on the other hand, were overtreated with tamoxifen in 31% of cases. In node-positive, premenopausal women compliance with guidelines was far better, with a 91–96% rate of chemotherapy prescription. In node-positive, postmenopausal, estrogen receptor-positive patients chemotherapy was unduly prescribed in as many as 56% of cases. Comparison of clinical practice with the next version of the guidelines (1998) showed a somewhat better compliance. Conclusions Despite the availability of official and authoritative guidelines, adjuvant treatment prescription for early breast cancer in Lombardy in 1997 was suboptimal, especially in well-defined subgroups of patients.