Crude and/or refined oil products accidentally released into coastal or oceanic waters will result in the formation of surface slicks that create hazards to sea life. Such releases can be mitigated by the use of chemical dispersants under appropriate conditions. However, the potential performance of these agents for dispersing oil in marine situations must be known prior to their use at spill sites. Therefore, reliable methods for evaluating and quantifying performance measurements for dispersants are necessary so that dispersant performance can be better estimated. An important component in any testing method for candidate dispersants (as well as other spill treating chemicals) is the precision associated with repeated measurements made with the method. This paper presents a statistical evaluation of results from tests that were conducted to evaluate the performance of different dispersants with a number of laboratory test methods. The overall test design for the study included three test methods, two to five test oils (depending on the method evaluated), three commercially available dispersants, and analytical measurements at three spectrophotometric wavelengths. The evaluation was conducted in support of the research program of EPA's Releases Control Branch, Risk Reduction Engineering Laboratory to support the EPA work group concerned with Subpart J (Dispersant Effectiveness and Toxicity) of the National Contingency Plan (NCP). A major goal of the parent study on which this work is based has been to estimate the variability or precision for repeated measurements with test methods and provide the best estimate of this precision for the proposed revision of Subpart J of the NCP for dispersant performance testing.