Aquatic macroinvertebrates were collected from two sections of Sink Branch, a central Florida stream disturbed by surface mining, channelization and grazing. Benthic organisms were sampled over a twelve month period with “stovepipe” substrate cores, drift nets, dip nets and multiple plate artificial substrate samplers. Species richness, organism density and Shannon-Weiner species diversity values were calculated for each sample. Sample variability was estimated by calculating coefficients of variation. With the exception of the dip nets which were considerably more diverse, none of the methods revealed significant differences in the diversity of samples. The four sampling devices did not census the number of species (species richness) equally well; the order of effectiveness was dip nets > drift nets > multiplates > substrate cores. There were no significant differences in variability of the species richness data among sampling techniques or among samples collected by the same technique. Organism density measured with substrate core and multiplate samplers was identical between sample types. Organism densities in the drift net, multiplate and core samples did not vary over time. A comparison of the coefficients of variation for the density data indicated that there were no significant differences in variability between sample techniques. Our results suggest that artificial substrate samplers may not reduce sampling variability in sandy Coastal Plain streams. In addition, the decision to use any of a variety of sampling techniques should be based on the nature of the assessment since sampling devices differ in the types of data they produce.