The rate of creatine kinase (CK) release from isolated frog extensor longus digiti IV ("toe") muscle increased on transfer from the control Ringer solution (referred to as 1.0 T) to a hyperosmotic solution (1.0 T plus 400 mM glycerol) or to hypertonic solutions (1.5-2.5 T) prepared by addition of NaCl or sucrose to 1.0 T. On return to 1.0 T, CK release rate was further elevated, reaching peak values (14.4- and 12-fold above the basal value, respectively, with the glycerol treatment and with solution 2.5 T) within 10-15 min and declining subsequently toward base line. Although the stimulation of CK release by changes in external tonicity increased with the magnitude of the osmotic step change, there was no enhancement of release when the tonicity changes were performed gradually in 0.1-T steps at 1-min intervals. The stimulation of CK release by tonicity changes cannot be ascribed to irreversible loss of functional integrity of the surface or tubular membranes or to CK diffusion through aqueous pores. The possibility that an exocytosislike process is involved is discussed.