Protein folding is a hierarchical process, driven by the accumulation of increments of free energy from local interactions between neighboring residues, secondary structural elements, domains and subunits. The latter represent independent folding units. Thus, the folding kinetics divide into the collapse of sub-domains and domains and their merging to form the compact tertiary fold. In proceeding to oligomeric proteins, docking of structured monomers is the final step. In agreement with this mechanism, in vitro experiments show that the overall mechanism of folding and association obeys uni-bimolecular kinetics with aggregation as a competing side reaction. In vivo, accessory proteins serve to shift the kinetic partitioning between assembly and misassembly toward the native state. So far, co- and post-translational protein folding in the cell has been withstanding a detailed kinetic analysis. Despite obvious differences between the crowded cytosol and optimized in vitro folding conditions, the general mechanism of protein self-organization within and without the cell seems to be similar. Effects of solvent parameters on the rate and mode of protein folding are less significant than predicted. Addition of small ligands and compatible solutes allow nucleation steps and viscosity effects to be analyzed. The absence of chimeras after synchronous in vitro reconstitution of oligomeric enzymes proves subunit interactions to be highly specific.