There is evidence that kinetics of tumour markers (TMs) CEA, CA15-3 and CA19-9 provide valuable information about disease state over time in patients with advanced breast and colorectal cancer but the literature contains differences in methodology so comparing findings is difficult. By modifying criteria developed by Rustin and colleagues [1-5] in ovarian carcinoma we have retrospectively identified a subset of patients (those with progressive (P) TMs) where survival is significantly reduced compared with those with responsive (R) TMs. This is true for CEA, CA15-3 and CA19-9 at the first chemotherapy given in advanced disease (chem1) (Hazard ratios (HR) = 9.99, 8.89, 5.75, P ¿ 0.001 in all cases) and CEA and CA19-9 at the second chemotherapy (chem2) (HR = 7.95, 9.00, P = 0.001 and 0.002 respectively) in patients with breast cancer. It is also true for CEA at chem1 in patients with colorectal cancer (HR = 2.51, P <0.001). Further studies are necessary to see if treatment directed by these criteria can influence survival. CEA and CA19-9 Rustin category in colorectal patients and CA15-3 Rustin category in breast patients correlated significantly with radiological category at chem1 and chem2 (CEA rs = 0.45 and 0.43, CA19-9 rs = 0.26 and 0.35, CA15-3 rs = 0.28 and 0.44). CA19-9 also correlates with radiological category at chem2 (rs = 0.38) in breast patients. This provides valuable information because RECIST criteria can delay radiological identification of disease progression compared with WHO criteria [6, 7] and new therapies may act to stabilise tumour growth rather than reduce it [8]. / Oncology Research Trust Fund at Airedale NHS Trust