Much uncertainty still exists regarding spatial and temporal variability of stable isotope ratios (13C/12C and D/H) in different CH4-emission sources. Such variability is especially prevalent in freshwater wetlands where a range of processes can influence stable isotope compositions, resulting in variations of up to approximately 50% for delta13C-CH4 and approximately 50% for deltaD-CH4 values. Within a temperate-zone bog and marsh situated in southwestern Ontario, Canada, gas bubbles in pond sediments exhibit only minor seasonal and spatial variation in delta13C-CH4, deltaD-CH4 and delta13C-CO2 values. In pond sediments, CO2 appears to be the main source of carbon during methanogenesis either directly via CO2 reduction or indirectly through dissimilation of autotrophic acetate. In contrast, CH4 production occurs primarily via acetate fermentation at shallow depths in peat soils adjacent to ponds at each wetland. At greater depths within soils, sigmaCO2 and H2O increasingly exert an influence on delta13C- and deltaD-CH4 values. Secondary alteration processes (e.g., methanotrophy or diffusive transport) are unlikely to be responsible for depth-related changes in stable isotope values of CH4. Recent models that attempt to predict deltaD-CH4 values in freshwater environments from D/H ratios in local precipitation do not adequately account for such changes with depth. Subenvironments (i.e., soil-forming and open water areas) in wetlands should be considered separately with respect to stable isotope signatures in CH4 emission models.