In this study, we apply a hybrid-resolution model, namely, PACE, to characterize the free energy surfaces (FESs) of Trp-cage and a WW-domain variant along with the respective folding mechanisms. Unbiased, independent simulations with PACE are found to achieve together multiple folding and unfolding events for both proteins, allowing us to perform network analysis of the FESs to identify folding pathways. PACE reproduces for both proteins expected complexity hidden in the folding FESs, in particular metastable non-native intermediates. Pathway analysis shows that some of these intermediates are, actually, on-pathway folding intermediates and that intermediates kinetically closest to the native states can be either critical on-pathway or off-pathway intermediates, depending on the protein. Apart from general insights into folding, specific folding mechanisms of the proteins are resolved. We find that Tip-cage folds via a dominant pathway in which hydrophobic collapse occurs before the N-terminal helix forms; full incorporation of Trp6 into the hydrophobic core takes place as the last step of folding, which, however, may not be the rate-limiting step. For the WW-domain variant studied, we observe two main folding pathways with opposite orders of formation of the two hairpins involved in the structure; for either pathway, formation of hairpin I is more likely to be the rate-limiting step. Altogether, our results suggest that PACE combined with network analysis is a computationally efficient and valuable tool for the study of protein folding.