To evaluate the relationship between obesity, analyzed by different indicators, and lung cancer incidence, literature search was conducted in the PubMed, Web of Science, EBSCO, Ovid, and China National Knowledge Infrastructure databases for articles published until December 2018. Twenty-eight prospective cohort studies were identified, with 28 784,269 participants and 127,161 lung cancer cases were included in the analysis. The combined relative risks (RRs) with 95% CIs for the highest versus normal category of body mass index (BMI) were RR = 0.77 (95% CI: 0.72-0.82), but the inverse association disappeared for never smokers or small cell carcinoma after stratifying the smoking status or histological cancer types, respectively. Further analysis considered lag time and excluded the effects of preclinical cancer, there is no statistically significant inverse association between BMI and lung cancer risk, RR = 0.89 (95% CI: 0.66-1.19). In contrast, the combined RRs with 95% CIs for the highest versus lowest category of waist circumference (WC) were RR = 1.26 (95% CI: 1.14-1.39). Therefore, due to multiple confounders existed, BMI might not be an appropriate indicator for obesity when study lung cancer risk. The significantly positive relationship between WC and lung cancer risk indicated there might have an etiological connection between central obesity and lung cancer development.