BACKGROUND: Macrophage inhibitory cytokine-1(MIC-1) is a potential modulator of systemic inflammation and nutritional depletion, both of which are adverse prognostic factors in oesophago-gastric cancer (OGC). METHODS: Plasma MIC-1, systemic inflammation (defined as plasma C-reactive protein (CRP) of X10 mg l -1 or modified Glasgow prognostic score (mGPS) of X1), and nutritional status were assessed in newly diagnosed OGC patients (n ¼ 293). Healthy volunteers (n ¼ 35) served as controls. RESULTS: MIC-1 was elevated in patients (median ¼ 1371 pg ml -1 ; range 141 -39 053) when compared with controls (median ¼ 377 pg ml -1 ; range 141 -3786; Po0.001). Patients with gastric tumours (median ¼ 1592 pg ml -1 ; range 141 -12 643) showed higher MIC-1 concentrations than patients with junctional (median ¼ 1337 pg ml -1 ; range 383 -39 053) and oesophageal tumours (median ¼ 1180 pg ml -1 ; range 258 -31 184; P ¼ 0.015). Patients showed a median weight loss of 6.4% (range 0.0 -33.4%), and 42% of patients had an mGPS of X1 or plasma CRP of X10 mg l -1 (median ¼ 9 mg l -1 ; range 1 -200). MIC-1 correlated positively with disease stage (r 2 ¼ 0.217; Po0.001), age (r 2 ¼ 0.332; Po0.001), CRP (r 2 ¼ 0.314; Po0.001), and mGPS (r 2 ¼ 0.336; Po0.001), and negatively with Karnofsky Performance Score (r 2 ¼ À0.269; Po0.001). However, although MIC-1 correlated weakly with dietary intake (r 2 ¼ 0.157; P ¼ 0.031), it did not correlate with weight loss, BMI, or anthropometry. Patients with MIC-1 levels in the upper quartile showed reduced survival (median ¼ 204 days; 95% CI 157 -251) when compared with patients with MIC-1 levels in the lower three quartiles (median ¼ 316 days; 95% CI 259 -373; P ¼ 0.036), but MIC-1 was not an independent prognostic indicator. CONCLUSIONS: There is no independent link between plasma MIC-1 levels and depleted nutritional status or survival in OGC.