The potential release of pollutants during the dredging process has in recent years come under the scrutiny of federal and state agencies. A simple laboratory test and mathematical model were sought to predict the concentration of contaminants released at the point of a dredging process. This research investigates the effect of initial suspended solids (SS) concentration, mixing time, and settling time in a 4-L graduate cylinder on the final water column concentrations of SS, PCBs, and selected metals (Cu, Pb, and Cd). The objective was to determine the conditions that could best simulate the actual contaminant concentrations measured in a pilot dredging operation of New Bedford Harbor. Results indicate that most of the PCBs found in the column after settling were sorbed onto suspended particles rather than dissolved. Aeration time had little effect on final SS or PCB concentration. Settling time had the most influence of the test parameters on SS and PCBs but beyond 1 hour, the effect was minimal. Analysis of settling rate indicated that flocculent rather than discrete settling occurred. From a matrix of laboratory testing using initial SS, aeration time, and settling time as variables, a non-linear regression model was used to predict final SS concentrations (from which PCBs released is calculated). Also the soluble PCB concentration was predicted by a simple equilibrium partitioning model. Particle size distributions (PSD) analyses were analyzed (Coulter Counter) on solids remaining in the laboratory column for various combinations of experimental conditions. PCB and SS concentrations obtained in New Bedford Harbor from the ports of the dredgehead and from the plume were compared to those found in the laboratory test. In general, the field and laboratory data were within one order of magnitude. However, the scatter in field data was considerable and this must be included in any analysis of predictive capability of a simple laboratory test and mathematical model. Finally, if sorbed PCB concentrations, initial suspended solids, and settling time are measured, then the total, particulate, and soluble PCBs in the water column can be predicted without elutriate tests by using the simple equilibrium partitioning model and the non-linear regression model developed to determine final SS.