The objective was to evaluate the effects of two (2×) vs three (3×) times per day milking on milk production and milk composition in dairy cows. Fourteen scientific papers, containing production data from 16 trials, where dairy cows were milked 2× or 3×, were analysed using meta-analysis with fixed and random-effects with the R statistical program. The degree of heterogeneity and publication bias were measured with the I2 statistic and Begg’s test, respectively. In addition, the meta-regression analysis explored other sources of heterogeneity for the response. The estimated effect size of 2× and 3× milkings was calculated for dry matter intake (DMI), milk production, and milk composition. Dry matter intake, milk production, and milk fat and protein yields showed substantial heterogeneity (I2>50%). Whereas milk fat-percentage had moderate heterogeneity (I2<50%), and milk protein had no (I2=0%)heterogeneity. The year of publication, trial duration, and cattle breed did not influence production response parameters to milking frequency. We found no evidence of publication bias for the parameters evaluated (Begg’s test; P>.05). Cows milked 2× produced less milk (2.23 kg/d), less milk fat (0.06kg/d), and less milk protein (0.05 kg/d). In contrast, the fat percentage was lower (0.07 units) in 3×, compared with 2× milking frequency. There was no effect of milking frequency on DMI and milk protein percentage. In conclusion, milk production and milk fat and protein yields improves as milking frequency increase from 2× to 3× daily, without affecting DMI. The implementation of 3× milking frequency must consider dairy cow management, labor, and milking parlour infrastructure, particular to each dairy farm.