1. Single dispersed cells obtained by collagenase treatment of longitudinal muscle of rabbit small intestine were voltage clamped with low-resistance patch pipettes and membrane current was measured. 2. In cells held at -20 or -30 mV, a discharge of spontaneous transient outward currents (STOCs) was usually seen; these are believed to represent the sporadic release of calcium from storage sites in the cell in relation to TEA-sensitive, 4 AP-resistant, calcium-activated potassium channels. 3. Caffeine (20 mM) externally applied, accelerated and then abolished STOCs; carbachol (0.1 mM) had similar effects; the initial burst of STOCs was often carried on a large, temporary, outward current which could occur alone. This was suggested to be caused by the rapid release of stored calcium in relation to calcium-activated potassium channels. 4. If STOCs were abolished by caffeine (or carbachol) then carbachol (or caffeine) did not evoke outward current indicating that these drugs act on the same calcium store but by different pathways. Inclusion of ryanodine (10(-8)-10(-4) M) in the patch pipette abolished STOCs soon after establishing whole-cell recording mode; afterwards, outward current to caffeine or to carbachol could not be evoked. 5. STOCs were quickly abolished in cells patched with pipettes filled with GTP gamma S (0.1-1 mM) or Gpp(NH)p (0.1-1 mM) but were large or normal in size in cells where GDP beta S (0.1-1 mM) was included in the pipette. GTP gamma S or Gpp(NH)p in the cell abolished outward current to caffeine or to carbachol, but had no effect on calcium-activated potassium channel activity in isolated patches or on a TEA-sensitive, 4-AP-resistant, outward potassium current evoked in single cells by stepping positively from a -20 mV holding potential. These results suggest that the effect of guanine nucleotide analogues are on the calcium store rather than on calcium-activated potassium channels. 6. The effects of GTP gamma S or Gpp(NH)p could be explained if they depleted calcium stores via a G-protein mechanism; this effect may involve activation of phospholipase C enzyme (PLC) and D-myo-inositol 1,4,5-trisphosphate (IP3) production as well as a direct effect on stores. However a separate G-protein-independent pathway of activation of PLC by muscarinic receptor activation may exist as calcium release by carbachol was large or normal in cells filled with GDP beta S.