Accurate prediction of methane (CH₄) emission from enteric fermentation is a key factor in assessing mitigation strategies on dairy farms. The aim of this study was to use dietary characteristics and animal performance data to predict CH₄ emissions from dairy cows. A database was built from measurements on individual cows from 6 experiments (56 Holstein cows, 18 treatments, 193 observations) assessing the impact of dietary modification on CH₄ production. Methane, in these trials, was determined using respiration chambers. Preliminary analyses with simple regression and Pearson correlation were performed to assess the relationship of animal variables (milk yield (MY), milk fat, milk protein, DIM, BW, DMI) and diet variables (OM, CP, NDF, ADF, Starch, EE and Rumen-inert fat: as % of DM) with CH₄ emissions and the potential collinearity between the independent variables. Backward multiple regression was then performed using the MIXED procedure of SAS 9.4 to predict CH₄ emissions with cow included as random effect. Five randomization with 5-fold cross validation were used to validate the equation. The equation with the best r (0.84) and the lowest RMSE (39.9) was: CH₄ (g/d) = −1,607.8 + 2.0 × MY (kg/d) + 56.6 × Milk Fat (%) + 11.8 × DMI (kg/d) + 0.2 × BW (kg) + 74.0 × NDF (% of DM) – 1.1 × NDF² (% of DM) + 2.6 × Starch (% of DM) + 14.3 × Rumen-inert fat (% of DM). Predicted CH₄ using the proposed equation was 474 ± 63 g of CH₄/cow/d, which agreed with the observed value (476 ± 73 g of CH₄/cow/d), which in turn was substantially lower than predicted by the IPCC equation (535 ± 59 g of CH₄/cow/d; r = 0.62 and RMSE = 57.0). Further analyses showed no mean bias (1.8 g of CH₄/cow/d; P = 0.53) or linear bias (−0.03; P = 0.49) for the proposed equation. In contrast, the IPCC equation exhibited a mean bias of −59.5 g of CH₄/cow/d (P = 0.001) and a linear bias of −0.23 (P = 0.001). The maximum bias with this equation (36 g of CH₄/cow/d) was however less than the prediction standard error (58 g of CH₄/cow/d). The proposed model is superior to IPCC equation to predict CH₄ emission and evaluate mitigation strategies on Canadian dairy farms.