Conventional treatment for brain metastases (BM) is whole-brain radiotherapy (WBRT). Efficacy is poor. It might be increased by a potent radiosensitiser such as gemcitabine which is believed to cross the disrupted blood -brain barrier. Primary objective of this study was to determine the maximum tolerated dose (MTD) of twice weekly gemcitabine given concurrently with WBRT. Patients with BM from carcinoma were included. The dose of WBRT was 30 Gys (10 daily fractions). Gemcitabine was given 2 -4 h prior to WBRT on days 1 and 8 for the first cohort of patients and then on days 1, 4, 8 and 11. Starting dose was 25 mg m À2 , escalated by 12.5 mg m À2 increments. At least three patients were included per level. Dose limiting toxicity (DLT) was defined as grade 4 haematological or grade X3 nonhaematological toxicity. A total of 25 patients were included; 74% had a PS 1 (ECOG). In all, 23 had non-small-cell lung cancer, six colorectal, four breast, two renal cell and one oesophageal carcinoma. A total of 92% had concurrent extracranial disease. Six had single BM, 13 had two or three BM and six multiple. Up to 50 mg m À2 (level 4) no DLT was observed. At 62.5 mg m À2 , one out of six patients developed DLT (thrombocytopenia-bleeding). The next dose level (75 mg m À2 ) was abandoned after grade 4 bone marrow toxicity (fatal neutropenic sepsis) was seen in one out of two patients. So that the dose of 50 mg m À2 will be taken forward for further study. British Journal of Cancer (2005) (Patchel, 2003). Except for very few cases where surgery may be indicated, the mainstream accepted therapeutic modality is whole brain radiotherapy (WBRT). Gamma-knife radiosurgery and conformal radiotherapy are recent developments that may confer a prognosis improvement to selected patients mostly in cases of single BM and controlled extracranial disease. The prognosis of the vast majority of patients who develop BM is poor. The median survival for treated patients with WBRT is approximately 4 months and the 1-year survival only 12% (Lagerwaard et al, 1999). Chemotherapy is not usually given and if the patient is actually receiving chemotherapy, it is often discontinued. This attitude has been fostered by the theoretical difficulty of drugs to pass the blood -brain barrier and achieve therapeutic significance in brain micrometastases. This however is probably not the case with macroscopic metastases. Brain metastases in patients with many chemosensitive tumours (e.g. testicular cancer, lymphoma, choriocarcinoma) are well known to respond to chemotherapy. A recent study by Postmus et al (2000) in patients with small-cell carcinoma and BM showed a 22% response rate to chemotherapy alone. However, the combination of chemotherapy and radiotherapy produced a 57% response rate and the control in the brain disease in these patients was longer than in patients who received only chemotherapy. A phase III study of early vs late WBRT with concurrent cisplatin and vinorelbine in patients with non-smallcell lung cancer (NSCLC) and BM was reported by Robi...