Even if the prognostic role of SUVmax of 18-FDG-PET has been largely investigated, many issues regarding its relationship with pathologic staging and histological subtypes still remain controversial. This retrospective study investigated the prognostic significance of SUVmax in 119 completely resected, pathologically proven NSCLC. The SUVmax values resulted significantly related to histological subtypes (p<0.001), histological grading (p<0.001), and pathologic stage (p<0.001). The optimal cut-off value of SUVmax to predict prognosis in the whole series was 6.7 (p=0.029). 2-Year disease-specific survival (DSS) was 91% for SUVmax < or =6.7 and 55% for SUVmax >6.7 (p<0.001). SUVmax still remain a significant predictor of survival in Stage IB (2-year DSS of 100% for SUVmax < or =6.7; 51% for SUVmax >6.7, p=0.016). The optimal cut-off values of SUVmax to predict prognosis were 5 for adenocarcinoma (p=0.027) and 10.7 for other non-adenocarcinoma NSCLC subtypes (p=0.010). These histologic-specific cut-offs resulted significantly related to survival when stratified for stage: 2-year DSS for Stage IB adenocarcinoma were 100% for SUV< or =5 and 40% for SUVmax >5 (p=0.051); 2-year DSS for Stage IB non-adenocarcinoma were 83% for SUVmax < or =10.7 and 26% for SUVmax >10.7 (p=0.018). Adenocarcinomas showed significantly lower survival results respect to other NSCLC for intermediate SUVmax level (range 5.5-11.3) (p=0.021). High SUVmax resulted an independent negative prognostic factor at multivariate analysis (HR of 15.7, 95% CI of 2.50-98.44, p=0.003). In conclusion, SUVmax represents a significant prognostic factor in surgically resected NSCLC but a great variability between different histological subtypes, even when adjusted for stage, is present and could be considered when planning future trials on prognostic role of FDG uptake.