Effect of chain architecture on the chain size, shape, and intrinsic viscosity was investigated by performing molecular dynamics simulations of polymer solutions in a good solvent. Four types of chains - linear, comb shaped, H-shaped, and star - were studied for this purpose using a model in which the solvent particles were considered explicitly. Results indicated that the chain length (N) dependence of the mean squared radius of gyration of the chains followed a power-law behavior 〈Rg2〉1/2∼Nυ with scaling exponents of υ = 0.605, 0.642, 0.602, and 0.608, for the linear, comb shaped, H-shaped, and star shaped chains, respectively. The simulation results for the geometrical shrinking factor were higher than the prior theoretical predictions for comb shaped chains. Analysis of chain shape demonstrated that the star chains were significantly smaller and more spherical than the others, while the comb and H-shaped polymer chains showed a more cylindrical shape. It is shown that the intrinsic viscosity of the chains can be calculated by plotting the specific viscosity determined from simulations against the solution concentration. The intrinsic viscosity exhibited linear behavior with the reciprocal of the overlap concentration for all chain architectures studied. The molecular weight dependence of the intrinsic viscosity followed the Mark-Houwink relation, [η] = KMa, for all chain architectures. When comparing the calculated values of exponent a with the literature experimental values, agreement was found only for the H and star chains, and a disagreement for the linear and comb chains. The viscosity shrinking factor of the branched chains was compared with the available experimental data and the theoretical predictions and a general agreement was found.