There are various sources of contamination that are known to occur. Each can be monitored by reference to staff elimination databases and negative controls in order to estimate or capture the effect of transfer from scientists and police investigators, reagent contamination, and cross-contamination from the concurrent processing of another sample. The current study showed how negative controls can be used to predict the performance of a DNA unit in terms of an estimate of the expected number of false-positive results, along with their associated likelihood ratios. In addition, it showed how the use of reporting guidelines affects the results. The description of the method addresses negative controls; how contamination arises and how the process can be monitored to produce predictive models of the consequences; the method for estimating Pc; the collection of data; and the details of the SIMLAB computer program, which produced the simulation model. The simulation model was used to combine results of negative controls with actual casework DNA profiles to assess the probability that laboratory contaminants will yield reportable profiles, along with their likelihood ratios. Using data from an operational DNA unit as an example, the simulation showed that the risk of contamination is linked to guidelines used to interpret DNA profiles. The simulation model predicted the levels of contamination that can be expected in the process and also how the process can be made more robust by changing reporting guidelines. The likeliest outcome of a contamination event is false exclusion rather than false inclusion. The probability of contamination is always greater than zero; however, the effect of contamination can be mitigated by comparing all casework profiles against staff databases, including manufacturers of plasticware; however, this can capture only a portion of the contamination events. It is possible to assess the prevalence of contamination and its effect by using expert systems that analyze negative controls. 4 tables, 3 figures, and 12 references