Vitamin E (VE) is broadly recognized by its effects on immune response and meat quality of broiler chickens. However, studies have presented controversial results of VE effects on broiler performance. A meta-analysis was conducted to estimate the linear effect of VE supplementation on body weight (BW, 39–49 d of age), daily weight gain (DWG), daily feed intake (DFI) and feed conversion ratio (FCR) of male broiler chickens, using a data set composed by 33 peer-reviewed journal articles from 2010 to 2014. Data were analyzed by linear mixed models approach, where studies (s) were assumed as random effect whereas VE total amount on diet as fixed effect. VE levels in the database were: mean = 91.89 ± 106.6 mg/kg, min = 0 mg/kg, max = 400 mg/kg, while the response variables presented the following averages and observations: BW = 2.16 ± 0.66 kg, n = 41, s = 12; DWG = 61.6 ± 16.02 g, n = 44, s = 13; DFI = 91.7 ± 19.6 g, n = 49, s = 13; FCR = 1.78 ± 0.18 g/g, n = 56, s = 16. The necessity to model error heteroscedasticity was evaluated based on Akaike’s information criterion and Schwarz criterion. Significance was declared at P ≤ 0.05. None of the analyzed variables was linearly influenced by VE supplementation, where the estimated slopes were not different from zero with P-values equal to: 0.99, 0.80, 0.22, 0.40, for BW, DWG, DFI and FCR, respectively. The absolute value of correlation between random terms (i.e., intercept and slope) was never greater than 0.27, what indicates that VE effect is independent of variables’ values range, with the exception of DFI (r = −0.86). Although 33 studies were evaluated in this research, few of them presented all performance variables simultaneously, what could have reduced the meta-regression power of analysis. Moreover, most studies presented a small range and few VE levels (i.e., less than 4 levels). More standardized and integrated research should be conducted to investigate possible effects of VE on broilers performance; nevertheless, there is not yet, in the current literature, any evidence indicating that the non-effect hypothesis should be rejected.