Residues 519-524 of Dictyostelium myosin II form a small surface loop on the actin binding face, and have been suggested to bind directly to actin through high affinity hydrophobic interactions. To test this hypothesis, we have characterized mutant myosins that lack this loop in vivo and in vitro. A mutant myosin in which this loop was replaced by an Ala residue (delta519-524/+A) was non-functional in vivo. Replacement with a single Gly residue instead of Ala yielded partial function, suggesting that structural flexibility, rather than hydrophobicity, is the key feature of the loop. The in vivo phenotype of the mutant enabled us to identify a number of additional amino acid changes that restore function to the delta519-524/+A mutation. Intriguingly, many of these, including L596S, were located at some distances away from the 519-524 loop. We have also isolated suppressors for the L596S mutant myosin, which was not functional in vivo. The suppressors for delta519-524/+A and those for L596S showed complementary charge patterns. In ATPase assays, delta519-524/+A S1 showed very low activity and little enhancement by actin, whereas L596S S1 was hyper active and displayed enhanced affinity for actin. In motility assays, delta519-524/+A myosin released actin filaments upon addition of ATP and was unable to support movements. L596S myosin was also inactive, but in this case actin filaments stayed immobile even after the addition of ATP. Transient kinetic measurements demonstrated that delta519-524/+A S1 is not only slower than wild type to bind actin filaments, but also slower to dissociate from actin filaments. Based on these results, we concluded that the 519-524 loop is not a major actin binding site but aids actin binding by facilitating a critical conformational change.