We have analyzed the effects of different components of the GroE chaperonin system on protein folding by using a nonpermissive substrate (i.e., one that has very low spontaneous refolding yield) for which rate data can be acquired. In the absence of GroES and nucleotides, the rate of GroEL-mediated refolding of heat- and DTT-denatured mitochondrial malate dehydrogenase was extremely low, but some three times higher than the spontaneous rate. This GroEL-mediated rate was increased 17-fold by saturating concentrations of ATP, 11-fold by ADP and GroES, and 465-fold by ATP and GroES. Optimal refolding activity was observed when the dissociation of GroES from the chaperonin complex was dramatically reduced. Although GroEL minichaperones were able to bind denatured mitochondrial malate dehydrogenase, they were ineffective in enhancing the refolding rate. The spectrum of mechanisms for GroE-mediated protein folding depends on the nature of the substrate. The minimal mechanism for permissive substrates (i.e., having significant yields of spontaneous refolding), requires only binding to the apical domain of GroEL. Slow folding rates of nonpermissive substrates are limited by the transitions between high- and low-affinity states of GroEL alone. The optimal mechanism, which requires holoGroEL, physiological amounts of GroES, and ATP hydrolysis, is necessary for the chaperonin-mediated folding of nonpermissive substrates at physiologically relevant rates under conditions in which retention of bound GroES prevents the premature release of aggregation-prone folding intermediates from the chaperonin complex. The different mechanisms are described in terms of the structural features of mini- and holo-chaperones.