Atlantic white cedar wetlands have pronounced hummock-hollow microtopography, in which the hummocks may be covered bySphagnum spp. or may have only a layer of needle and leaf litter present. I tested the hypotheses that characteristics of the hummock peat, including moisture content, bulk density, redox potential and fiber content, would vary linearly with elevation above the water table, and that these relationships would not vary between moss- and litter-covered surfaces. These hypotheses underlie the frequent use of microsite elevation as an indicator of habitat differentiation in wetlands. I found that moisture and redox potential varied with elevation, but bulk density and fiber content were independent of elevation. Furthermore, moisture and redox status onSphagnum-covered microsites had a weaker relationship to elevation (lower slope coefficient and lower r2) than on litter-covered microsites. Principal components analyses showed thatSphagnum-covered microsites were differentiated primarily on the basis of the relative content of undecomposed and decomposed fibers, while litter-covered sites were differentiated primarily on the basis of moisture and elevation. These results suggest that environmental gradients on hummock-hollow systems are complex and not a simple reflection of elevation. Environmental disturbances that reduceSphagnum cover may affect vegetation by creating more drought-prone microsites on the tops of the hummocks, and this may be important for management of cedar regeneration.