Protein folding in vivo is extremely intricate and challenging to examine or predict because the conformational changes, including folding, misfolding, and aggregation, are largely influenced by the cellular environment. Traditionally, cellular protein folding has been considered predominantly in the context of the Anfinsen postulate and molecular chaperones. However, accumulating evidence reveals that these models have limitations. In this review we revisit these models, and discuss co-translational folding, binding partner-mediated folding, and RNA-mediated folding as alternative or supplementary folding helpers. In addition, we discuss the folding helper systems mediated by macromolecules (e.g., ribosomes, membranes, and prefolded domains in multidomain proteins) that are tightly linked to newly synthesized polypeptides during protein biogenesis. These cis-acting folding helper systems, conceptually different from the trans-acting molecular chaperones, could play a crucial role in protein folding in vivo. Importantly, there is increasing evidence that the surface charges and excluded volume of macromolecules are important factors for stabilizing their connected polypeptides against aggregation. This stabilizing mechanism suggests that macromolecules including RNAs and proteins, let alone molecular chaperones, have an intrinsic ability to exert chaperoning function on their connected polypeptides independent of the linkage type between them. As an effective way to overcome the adverse effect of macromolecular crowding on protein folding, here we suggest that nascent polypeptide chains utilize the crowded environment in favor of productive folding by interacting with macromolecules.