A microkinetic model for CO methanation on nickel based on CO dissociation and stepwise hydrogenation of surface carbon is presented. Very good agreement with the methanation rates measured on Ni(100) by Goodman et al. (J. Catal. 63, 226 (1980)) and on nickel foils by Polizzotti and Schwarz (J. Catal. 77, 1 (1982)) is obtained by assuming that the rate-controlling step is hydrogenation of surface methylidyne and by treating the coverage of reactive carbon, theta(C), as a parameter. For not too low H-2:CO ratios it seems to be a good approximation to keep theta(C) constant. The validity of this approximation may be related to the observations that only a small part of the surface carbon is reactive and that the rate depends on previous treatments of the nickel surface supposed to create special sites. It is shown that previously published macrokinetic models, which can be tested, do not give rates in agreement with the single crystal and foil results. Analysis of the present model shows that the overall activation energy, E(a), of the reaction is mainly determined by the chemisorption energy of CO in the temperature range where E(a) is almost constant. The order of the H-2 dependence derived from the kinetic model is close to 1 in the entire range of reaction conditions considered, while the order of the CO dependence is - 1 at low temperatures increasing at the highest temperatures to between -0.12 and -0.06 depending on the total pressure.