Sediment quality assessment values were developed using a weight of evidence approach in which matching biological and chemical data from numerous modelling, laboratory, and field studies performed on freshwater sediments were compiled and analyzed. Two assessment values (a threshold effect level (TEL) and a probable effect level(PEL)) were derived for 23 substances, including eight trace metals, six individual polycyclic aromatic hydrocarbons (PAHs), total polychlorinated biphenyls (PCBs), and eight pesticides. The two values defined three ranges of chemical concentrations; those that were (1) rarely, (2) occasionally, and (3) frequently associated with adverse biological effects. An evaluation of the percent incidence of adverse biological effects within the three concentration ranges indicated that the reliability of the TELs (i.e., the degree to which the TELs represent concentrations within the data set below which adverse effects rarely occur) was consistently good. However, this preliminary evaluation indicated that most of the PELs were less reliable (i.e., they did not adequately represent concentrations within the data set above which adverse effects frequently occur). Nonetheless, these values were often comparable to other biological effects-based assessment values (which were themselves reliable), which increased the level of confidence that could be placed in our values. This method is being used as a basis for developing national sediment quality guidelines for freshwater systems in Canada and sediment effect concentrations as part of the Assessment and Remediation of Contaminated Sediments (ARCS) program in the Great Lakes.