Diagenesis of carbon, oxygen, nitrogen, and manganese at three sites on the Canadian continental margin is quantitatively compared and contrasted using results from a computer code (CANDI) published by Boudreau (1996a). The data at Station 3 (Cabot Strait) are well explained by the steady state output from CANDI, assuming a porewater balance created by diffusion and reaction only, whereas the data from Stations 4 (Emerald Basin-Scotia Shelf) and 5 (Scotia Slope) are not consistent, in one way or another, with this simple model. The deviations between model and data at Station 4 are best explained by nonsteady-state diagenesis. Model fits to the Station 5 ΣCO2 observations are improved dramatically by adding some irrigation at this site, but the ΣNH3 distribution appears to be subject to an additional anomalous transport to the O2 zone and subsequent oxidation to NO-3. The mechanism for this latter phenomena is unknown and in need of future research. In addition, the O2 and ΣCO2 profiles at all sites require the existence of at least two reactive organic matter types; furthermore, the initial amounts of these OM types at each station is strongly dependent on the intensity of particle bioturbation. Ammonia is preferentially regenerated at Station 3 at a high ratio of about 25 N to 106 C. The net kinetics of the deeper removal of Mn2+ appear to be fractional-order with respect to the concentration of this species, suggesting multiple removal processes. Finally, an oxidant balance, assuming steady state, indicates a considerable difference in the use of oxidants at each station even though the O2 fluxes are similar.