In order to determine whether angiogenesis is a prognostic marker in lung cancer, we performed a systematic review of the literature to assess the prognostic value on survival of microvessel count in patients with lung cancer. Published studies were identified by an electronic search in order to aggregate survival results, after a methodological assessment using a quality scale designed by the European Lung Cancer Working Party. To be eligible, a study had to deal with microvessel count assessment in lung cancer patients on the primary site and to provide survival analysis according to microvessel count expression. Microvessel count has been assessed on surgical samples by immunohistochemistry using factor VIII in 14 studies, CD34 in 10 and CD31 in eight. Respectively 1866, 1440 and 1093 non-small cell lung cancer patients were considered. The overall median quality scores were respectively 52, 59 and 59% for studies assessing microvessel count via factor VIII, CD34 and CD31, without significant difference between studies evaluable or not for meta-analysis nor between studies with significant or non significant results. Seven 'factor VIII' studies, nine 'CD34' and seven 'CD31' provided sufficient data allowing a metaanalysis on survival and were evaluable for results aggregation. This showed that a high microvessel count in the primitive lung tumour was a statistically significant poor prognostic factor for survival in non small cell lung cancer whatever it was assessed by factor VIII (HR: 1.81; 95% CI: 1.16 -2.84), CD34 (HR: 1.99; 95% CI: 1.53 -2.58) or CD31 (HR: 1.80; 95% CI: 1.10 -2.96). Variations in survival among the individual studies can be explained in addition to patients selection criteria by the heterogeneous methodologies used to stain and count microvessels: different antibody clones, identification of 'hotspots', Weidner or Chalkey counting method, cut-off selection. Microvessel count, reflecting the angiogenesis, appears to be a poor prognostic factor for survival in surgically treated non small cell lung cancer but standardisation of angiogenesis assessment by the microvessel count is necessary.