Free cytosolic calcium ([Ca2+]i) levels increase after the stimulation of either human basophils or mast cells with anti-IgE antibody. Previous studies found that the mast cell [Ca2+]i response was graded in magnitude, according to stimulus strength, and that the activated level was free of oscillations. The current studies demonstrate several new features of the mast cell and basophil response. First, in mast cells, the transition to activated [Ca2+]i levels was abrupt (width = 6.5 +/- 2 s), after a period of quiescence whose duration (5 to 300 s) was a function of the strength of the stimulus. At optimal concentrations of anti-IgE, 97% of mast cells showed only abrupt transitions and oscillation-free activated calcium levels. In contrast, basophils showed marked oscillations whose magnitudes were partially dependent on the strength of stimulus. Like the mast cell, there was a quiescent period before the first transition and this period was also dependent on the strength of the stimulus. Oscillations were generally superimposed on an elevated [Ca2+]i level at a frequency of 0.5 to 3/min, had half-widths of 5 to 20 s, and were markedly chaotic in the frequency domain. In general, oscillations were more apparent at suboptimal concentrations of anti-IgE. Despite the apparent contrast in basophil and mast cell responses, oscillations (1 or 2 in a 10-min period) could be observed in a small percentage of mast cells and some basophils showed characteristics of mast cells. We tentatively conclude that mast cells and basophils utilize a similar mechanism of calcium mobilization but that the nonlinear characteristics of the calcium response may account for the mast cell/basophil differences. These studies indicated that the calcium kinetics, as measured by population averages, did not reflect the kinetics observed at the single cell level. Both mast cells and basophils had characteristics which could be described as graded and characteristics resembling all-or-nothing processes; the magnitude of a response was graded according the strength of stimulus while the kinetics profile appeared as an all-or-nothing event.