A problem area in determining passive sampler reliability for the monitoring of hazardous substances was addressed. The author points out that passive sampling devices rely on diffusion to provide a flux of material into the sampler. The question of accuracy arises because diffusion is generally a slow process. Potential problems are encountered in the diffusive sampling of time dependent concentrations using surfaces which are comprised of an imperfect sink sorbent. Excessive errors can occur in time weighted average (TWA) concentration estimates unless it is certain that the concentrations of vapors to be monitored remain sufficiently constant over time. Depending on the time of a pulse occurrence, a considerable variety of net sampled masses is possible, even though the TWA concentration is fixed. Evidence was presented which indicated that this error can be significantly reduced through the use of several sorbent surfaces in tandem. Calculations were also given which showed that the remaining errors are analyzable. An example of a monitoring situation with a single sorbent, double quantity of the single sorbent, or addition of another adsorbent surface yielded respective error limits of -30 to +10 percent, -18 to +5 percent, and -5 to +3 percent respectively. The author indicates that these findings can be used in the design of samplers, to ensure that a proposed sampler is able to obtain whatever degree of accuracy is required. The author also notes that knowledge of worst case sampling conditions has application in the development of performance guidelines and reasonable tests for sampling method validation.