How the intracellular signal transduction machinery integrates numerous input signals to determine a cell's fate is an important yet unresolved question in cell biology. Since cell fate decisions involve switch-like selection from a finite number of discrete options, much effort has been focused on understanding the behavior of switch-like signal transduction systems. Experimental investigation of signal transduction traditionally relies on cell population average response measurements, such as immunoblotting. However, the inevitability of cell response heterogeneity makes population averaging inadequate for investigating switch-like systems, since averaging of heterogeneous “on” and “off” responses will mask switch-like behavior as analog. Thus, single cell measurements are needed to reveal switch-like behavior. In the current work, we use flow cytometry to obtain single cell measurements of Epidermal Growth Factor (EGF) induced ERK1/2 activation in Human Embryonic Kidney (HEK) 293 cells, and develop probabilistic models to characterize the resulting population response distributions. Through experimental and theoretical analyses we find that the ERK1/2 response distributions are well characterized by a convoluted Gamma-Gaussian distribution. Upon fitting the ERK1/2 responses to this distribution, we find that the response of ERK1/2 appears analog at short times and digital at longer times, suggesting that EGF causes a restructuring of the ERK1/2 signal response properties. Overall, applying the probabilistic modeling and analysis techniques presented in this study allow one to gain quantitative insight about switch-like signal transduction systems from single-cell measurements.