This research analyzed sediment contamination concentrations for mercury and lead in Lakes Ontario and Erie using a GIS-based kriging approach. Environment Canada provided sediment survey data for Lake Ontario (1968 and 1998) and Lake Erie (1971 and 1997/98). Collation and mapping of point measurement data without the application of interpolation methods does not allow for spatial data trends to be fully analyzed. The kriging technique enables the creation of interpolated prediction surfaces, with the advantage that the results can be statistically validated. Although data normality is not required, the kriging results for the historical datasets suggest that it may be desirable, as statistical validity was reduced due to some individual stations having very high contaminant concentrations. Three of the four models developed for the 1997/98 data were statistically valid. For both lakes, the more recent data reveal reduced concentrations of mercury and lead, and there has been an overall reduction in contamination levels. However, sediments in some areas still exceeded Canadian sediment quality guidelines. The areas of greatest sediment contamination in Lake Ontario were within the major depositional basins, presumably as a result of historical industrial activities in watersheds along the southern and western shoreline including the Niagara River. In Lake Erie, areas of greatest sediment contamination continue to be located in the western and south central portions of the lake in proximity to the Detroit River and major urban/industrial centres.