Serum carbohydrate antigen 19-9 (CA 19-9) has been identified as a useful tumour marker for diagnosis of exocrine pancreatic carcinoma, but its value for evaluating the response to chemotherapy with gemcitabine is not clear. Tumour regression in pancreatic carcinoma is hard to determine due to massive desmoplastic tissue. Furthermore, objective tumour response does not automatically transcribe into better survival. Therefore, clinical benefit response, a composed parameter consisting of factors like performance status, pain, and body weight was integrated in evaluating tumour response. The aim of this prospective study was to evaluate the usefulness of serial CA 19-9 measurements as a biochemical response marker and an outcome prognostic parameter in patients with advanced pancreatic cancer receiving gemcitabine treatment. A total of 46 consecutive patients (median age 66 years) suffering from histologically proven locally advanced or metastatic adenocarcinoma of the exocrine pancreas were analysed. Gemcitabine was applied for a median of 23 courses (range 6-76). Two patients achieved an objective complete remission, five an objective partial remission (overall response, OR=15.2%), while objective stable disease was documented in 19 and objective progressive disease in 20 patients. Patients with a decrease of >20% of the baseline CA 19-9 level after 8 weeks of chemotherapy had a significantly better median survival than patients with a rise or a decline <20%. The response of CA 19-9 >20% during chemotherapy was the only independent predictor of survival in a multivariate analyses. In contrast, neither objective tumour response nor clinical benefit response showed this level of significance. In conclusion, kinetics of CA19-9 serum concentration serves as an early indicator of response to gemcitabine chemotherapy in advanced pancreatic cancer.