The purpose of this study was to determine the maximum tolerated dose (MTD) of gemcitabine when given concurrently with standard radiotherapy for the treatment of chest wall recurrences, and to compare actuarial rates of local-regional control with those achieved in historical controls. Patients with unresectable chest wall recurrences were enrolled in a phase I trial of concurrent gemcitabine and radiotherapy. Gemcitabine was increased at 150 mg/m(2)/week increments, starting at 300 mg/m(2)/week. Radiotherapy was delivered to the chest wall and regional nodes to a total of 60 to 70 Gy in 2 Gy daily fractions. Treatment toxicity was assessed and a comparison of treatment outcome was performed between study patients and historical groups treated with either radiotherapy alone or excision followed by radiotherapy. The dose-limiting toxicities of neutropenia and thrombocytopenia occurred at the second planned dose of 450 mg/m(2)/week after accrual of only six patients, resulting in a MTD of 300 mg/m(2)/week. Myelosuppression and skin desquamation were commonly observed. Actuarial rates of local-regional control were 100%, 50%, and 90% at 2 years for the gemcitabine with radiotherapy, radiotherapy alone, and excision followed by radiotherapy groups, respectively (p = 0.105). The difference among the Kaplan-Meier curves for overall local-regional control was statistically significant at p = 0.007 in favor of combined gemcitabine and radiotherapy. The MTD of gemcitabine is 300 mg/m(2)/week when gemcitabine is delivered concurrently with radiotherapy for unresectable chest wall failures. This novel approach suggests excellent local-regional control when compared to historical controls. A phase II trial is warranted.