AbstractEvidence for the association of ‘soluble’ enzymes in vivo is extensive and compelling. These associations occur in all compartments of the cell of both prokaryotes and eukaryotes. Several factors present in vivo promote these associations among enzymes whose association in vitro is often too weak to detect. Several physiological advantages of the associated enzyme complexes can be identified, most (but not all) of which are the consequence of microcompartmentation of metabolites (substrate channeling). Substrate channeling of intermediates by either a ‘direct transfer’ process or ‘proximity effects’ can occur. The latter mechanism does not require the special molecular features needed for the direct transfer mechanism and may, therefore, exist in more general situations in the cell. Criticisms of these views are discussed. We argue that these criticisms have been largely answered by experiment and theory in recent years. Studies on simple systems in vitro, nevertheless, contribute important insights concerning the more complex phenomena in vivo.