The objectives were to use meta-analytic methods to determine the effects of amount of supplemental choline ion as rumen-protected choline (RPC) starting prepartum on production and health of dairy cows. The literature was systematically reviewed and 21 experiments, with up to 66 treatment means and 1,313 prepartum parous cows, were included. All experiments had a treatment with no supplemental choline (0 g/d; n = 30 treatment means), and the amount of choline ion supplemented to treated cows ranged from 5.6 to 25.2 g/d (n = 36 treatment means). Duration of pre- and postpartum feeding of RPC averaged (±standard deviation) 22.0 ± 6.0 and 57.5 ± 42.2 d, respectively. Data collected included the ingredient composition and chemical analyses of pre- and postpartum diets, amount of choline ion supplemented, number of cows per treatment, frequency of health events, and the least squares means and respective standard error of the means for production responses, liver composition, and blood parameters. The concentrations of net energy for lactation and metabolizable amino acids and protein (MP) in pre- and postpartum diets were predicted for each treatment mean using National Research Council (2001). Mixed model meta-analysis was conducted including the random effect of experiment and weighting by the inverse of the standard error of the means squared. Increasing supplementation of choline ion during transition linearly increased pre- (β = 0.0184 ± 0.00425) and postpartum dry matter intake (β = 0.0378 ± 0.00974), and yields of milk (β = 0.436 ± 0.112), energy-corrected milk (ECM; β = 0.422 ± 0.0992), fat (β = 0.00555 ± 0.000793), and protein (β = 0.0138 ± 0.00378). Nevertheless, an interaction between choline and postpartum metabolizable methionine as a percent of MP (METMPPo) was observed for yields of milk, ECM, and protein because as METMPPo increased, the positive response to choline on yields of milk, ECM, and protein decreased. Supplementing choline during transition tended to reduce the risks of retained placenta and mastitis, but it had no effect on metritis, milk fever, displaced abomasum and ketosis, or the concentration of triacylglycerol in the hepatic tissue postpartum. The median amount of choline ion supplemented was 12.9 g/d and responses in postpartum dry matter intake and yields of milk, ECM, fat, and protein to that amount of supplementation were 0.5, 1.6, 1.7, 0.07, and 0.05 kg/d, respectively. No interactions were observed between supplemental choline and prepartum dietary net energy for lactation or metabolizable methionine as a percent of MP. Collectively, feeding RPC during the transition period improves performance in parous cows. Increases in yields of milk and milk components were observed in spite of pre- and postpartum diets, although the increments in milk, ECM, and protein yields with supplementing choline decreased as the concentration of methionine in postpartum diets increased. The optimum dose of choline ion was not detected, but likely it is more than the 12.9 g/d fed in most experiments evaluated in the current meta-analysis. Finally, the meta-analysis identified lack of sufficient data to understand the role of supplemental choline in nulliparous cows.