The recent experimental data on stages and kinetic intermediates in protein folding are reviewed. It is emphasized that these data are consistent with the 'framework model' proposed by the author in 1973. The model implies that protein folds by stage mechanism (secondary structure - molten) globule state - native state) in such a way that the results of previous stages are not reconsidered in subsequent ones. Arguments are presented that both these hypotheses and available experimental data do not contradict the assumption that native structures of at least small proteins are nevertheless under thermodynamic rather than kinetic control i.e. correspond to global minima of free energy.