Chemoattractants added to cells of the cellular slime mold dictyostelium discoideum induce a transient elevation of cyclic GMP levels, with a maximum at 10 s and a recovery of basal levels at approximately 25 s after stimulation. We analyzed the kinetics of an intracellular cGMP binding protein in vitro and in vivo. The cyclic GMP binding protein in vitro at 0 degrees C can be described by its kinetic constants K(1)=2.5 x 10(6) M(- 1)s(-1), k(-1)=3.5 x 10(-3)s(-1), K(d)=1.4 x 10(-9) M, and 3,000 binding sites/cell. In computer simulation experiments the occupancy of the cGMP binding protein was calculated under nonequilibrium conditions by making use of the kinetic constants of the binding protein and of the shape of the cGMP accumulations. These experiments show that under nonequilibrium conditions by making use of the kinetic constants of the binding protein and the shape of the cGMP accumulations. These experiments show that under nonequilibrium conditions the affinity of the binding protein for cGMP is determined by the rate constant of association (k(1)) and not by the dissociation constant (k(d)). Experiments in vivo were performed by stimulation of aggregative cells with the chemoattractant cAMP, which results in a transient cGMP accumulation. At different times after stimulation with various cAMP concentrations, the cells were homogenized and immediately thereafter the number of binding proteins which were not occupied with native cGMP were determined. The results of these experiments in vivo are in good agreement with the results of the computer experiments. This may indicate that: (a) The cGMP binding protein in vivo at 22 degrees C can be described by its kinetic constants: K(1)=4x10(6)M(-1)s(-1) and K(-1)=6x10(-3)s(-1). (b) Binding the cGMP to its binding protein is transient with a maximum at about 20-30 s after chemotactic stimulation, followed by a decay to basal levels, with a half-life of approximately 2 min. (c) The cGMP to its binding proteins get half maximally occupied at a cGMP accumulation of delta[cGMP](10)=2x10(-8) M, which corresponds to an extracellular stimulation of aggregative cells by 10(-10) M cAMP. (d) Since the mean basal cGMP concentration is approximately 2x10(-7) M, the small increase of cGMP cannot be detected accurately. Therefore the absence of a measurable cGMP accumulation does not argue against a cGMP function. (e) There may exist two compartments of cGMP: one contains almost all the cGMP of unstimulated cells, and the other contains cGMP binding proteins and the cGMP which accumulates after chemotactic stimulation. (f) From the kinetics of binding, the cellular responses to the chemoattractant can be divided into two classes: responses which can be mediated by this binding protein (such as light scattering, proton extrusion, PDE induction, and chemotaxis) and responses which cannot be (solely) mediated by this binding protein such as rlay, refractoriness, phospholipids methylation, and protein methylation.