1. Phosphoribulokinase was purified 286-fold from extracts of autotrophically grown cells. 2. The enzyme had a molecular weight of 237000 and showed a pH optimum of 9.0 in both crude extracts and purified preparation. MgCl(2) was required for activity; full activation was obtained at 5mm-MgCl(2) and the K(m) was approx. 0.5mm. 3. The ATP-saturation curve was sigmoidal and the degree of positive co-operativity increased at higher MgCl(2) concentrations. The ATP-binding sites appeared to be non-interacting at low ribulose 5-phosphate concentrations. 4. Lineweaver-Burk plots for ribulose 5-phosphate showed abrupt transitions between apparently linear sections. The apparent K(m) and V(max.) values increased with increasing concentrations of ribulose phosphate. The transitions may be explained by a sequence of negative and positive co-operativity in the catalytic rate constants. 5. Phosphoribulokinase activity was inhibited by AMP and phosphoenolpyruvate and was activated by NADH. The presence of AMP or phosphoenolpyruvate increased s(0.5) (substrate concentration required for half-maximal velocity) for both ribulose 5-phosphate and ATP but V(max.) was not changed. The sigmoidicity of the ATP-saturation curve increased in the presence of AMP but was not affected by phosphoenolpyruvate. The transitions in the ribulose 5-phosphate-saturation curves were more abrupt in the presence of either inhibitor. NADH lowered the s(0.5) for both ribulose 5-phosphate and ATP. The activator did not affect the degree of positive co-operativity between ATP-binding sites, but the ribulose 5-phosphate-binding sites appeared to be non-interacting in its presence. 6. A sequence of positive and negative co-operativity in the interactions of AMP-binding sites was suggested by the Hill plots. In the presence of NADH (and phosphoenolpyruvate) the sensitivity to inhibition by AMP was less below a certain AMP concentration and increased above that concentration. 7. Examination of the interactions between ligands indicated that phosphoribulokinase can be regulated effectively by changes in effector concentrations similar to those reported to occur in vivo.