We examined the spatial structure of noise in optical recordings made with two commonly used voltage-sensitive dyes (RH795 and RH1691) in mouse barrel cortex in vivo, and determined that the signal-to-noise ratio of the two dyes was comparable when averaging over barrel-sized areas, or at single pixels distant from large blood vessels. We examined the spatiotemporal development of whisker- and electrically-evoked optical responses by quantifying the area of activated cortical surface as a function of time. Whisker and electrical stimuli activated cortical areas between 0.2-2.0 mm(2) depending on intensity. More importantly, both types of activation recruited cortical area at similar rates and showed a linear relationship between the maximal activated area and the peak rate of increase of the activated area. We propose a general rule of supragranular cortical activation in which the initial spreading speed of the response determines the total activated area, independent of the type of activation. Finally, despite comparable single-response kinetics, we observed greater paired-pulse depression of whisker-evoked responses relative to electrically-evoked responses.