A previously reported potential energy surface (PES) and a new barrierless PES (both based on ab initio data and describing the CH3 group as a pseudoatom) were used to study the O(D-1)+CH4--> OH+CH3 reaction with the quasiclassical trajectory (QCT) method. The new PES accurately reproduces the experimental rate constant values, in contrast to the previous PES. The QCT study was mainly performed at the relative translational energy (E-T) resulting from the photodissociation of N2O at 193 nm ([E-T]=0.403 eV), although the collision energy obtained from the photodissociation of O-3 at 248 nm ([E-T]=0.212 eV) was also considered. Good agreement between theory and experiment was obtained for the OH vibrational populations and for the OH rotational populations for the upsilon (')greater than or equal to2 vibrational levels, while the rotational distributions for upsilon (')=0-1 are more excited than in the experiment. The QCT results at E-T=0.403 eV satisfactorily reproduce the experimental kk(') angular distribution of the state-specific channel OH(upsilon (')=4, N-'=8) and the corresponding E-T(') distribution. For OH(upsilon (')=0, N-'=5) the reproduction of these properties is poorer, especially for the E-T(') distribution. At 0.403 eV the contribution of the abstraction mechanism to the reaction mode is negligible and two insertion like mechanisms (with fast or slow elimination) are found to be predominant, as suggested experimentally. The discrepancies observed between the QCT and experimental results can be explained on the basis of the defective description of the insertion/slow elimination mechanism provided by the model.