We compare independent data sets of the maximum stomatal conductance (g(smax), for single leaves) and bulk surface conductance (G(smax), for a vegetated surface including the plant canopy and soil) for evaporation. Data were obtained from field measurements, restricted to periods with plentiful soil water, adequate light, high relative humidity and moderate temperature. The data encompass most major vegetation types and a wide range of leaf area index (Lambda). Observed G(smax) is not systematically dependent on Lambda; and takes average values of 20 and 33 mm s(-1) for natural vegetation and agricultural crops. A similar pattern exists in the g(smax) data, which yield remarkably consistent average values of 6 and 12 mm s(-1), respectively, for natural vegetation and crops. Overall, the ratio G(smax)/g(smax) is consistently close to 3, for seven major vegetation types of diverse structure. A simple model accounts for the close relationship between g(smax) and G(smax), and in particular how G(smax) is conservative against Lambda because of the compensating decrease in plant canopy evaporation and increase in soil evaporation as Lambda diminishes. The results are important for development of parameters for biosphere-atmosphere interactions in models.