This study investigates both the methods and the multispecies and functional level effects of a release of a complex hydrocarbon mixture from sediments using a 60-d modified mixed flask culture (MFC) microcosm. Neat Jet-A was injected and mixed into the sediment with one nondosed and three dosed concentrations. Univariate and multivariate statistical and graphical techniques were used to detect patterns in the data. A slow release of the test material from the spiked sediment layer was obtained, and constituents of Jet-A were detected. Functional parameters (such as pH) were generally better at determining treatment groups than structural parameters (population densities). Analysis of the ability of the various parameters to detect treatment differences confirms that there is not one best indicator for the status of an ecological structure. Transient but statistically significant outcomes were seen at initial treatment concentrations as low as 2 microl/L. The higher concentration treatment groups could be identified as distinguished from nondosed or lower treatments at the end of the 63-d experiment. Each of the three multivariate techniques differed in their ability to distinguish treatment groups during the course of the experiment.