The mean-square radius of gyration (S-2) and intrinsic viscosity [eta] were determined for atactic poly(methyl methacrylate) (a-PMMA) in acetone at 25.0 degrees C, in nitroethane at 30.0 degrees C, and in chloroform at 25.0 degrees C in the range of weight-average molecular weight M(W) from 6.02 X 10(2) to 1.58 x 10(6). The present results for the gyration- and viscosity-radius expansion factors alpha(S) and alpha(eta) for a-PMMA along with the previous ones for atactic polystyrene (a-PS) and polyisobutylene are found to become functions only of the scaled excluded-volume parameter (z) over tilde defined in the Yamakawa-Stockmayer-Shimada (YSS) theory based on the helical wormlike chain model. Here, alpha(eta) for a-PMMA has been calculated by taking account of the dependence on solvent of the Flory-Fox factor phi(infinity), (or Phi(theta) defined in this paper). It is shown that ((z) over tilde) may be quantitatively explained by the YSS theory, while alpha(eta)((z) over tilde) may be approximately explained by the Barrett equation for alpha(eta) with (z) over tilde in place of the conventional excluded-volume parameter z : Thus the present results confirm that the quasi-two-parameter scheme is valid for both alpha(S) and alpha(eta) irrespective of the large differences in chain stiffness, local chain conformation, and solvent condition, indicating that there is no draining effect in alpha(eta) at least for the present systems. It is found that the effect of chain stiffness on alpha(S) and alpha(eta) for a-PMMA remains large even for M(W) > 10(6) as in the case of a-PS. The unperturbed values (S-2)(0) of (S-2) for a-PMMA in acetone at 25.0 degrees C, in acetonitrile at 44.0 degrees C (theta), and in n-butyl chloride at 40.8 degrees C (theta) are found to coincide with each other, indicating that (S-2)(0) Of a-PMMA is independent of solvent in contradiction to the prevailing view. Its dependence on solvent for a-PMMA that has been exclusively obtained so far from the Stockmayer-Fixman plot of [eta] is only apparent and the dependence on solvent of [eta](theta) must be regarded as arising from that of phi(infinity).