Soils adhering to shoes, car tires, and plant roots were compared with multiple control samples using dry color, ashed color, pH, saccharide levels, and particle size analysis. Color and pH were found to be as reproducible under simulated scenes of crime conditions as for true duplicates from homogeneous soils. The variation in saccharide levels was slightly greater than for true duplicates. An interpretational scheme developed for the silt and sand fractions of homogeneous soils was then applied to the samples from each simulated crime scene. Every comparison was found to be similar at the 99 percent confidence limit. Therefore, it is suggested that under normal circumstances during crime scene investigation, six control samples of soil should be taken with a dessert spoon and submitted for laboratory analysis. However, yet the significance of a similarity between two soil samples can only be established if data banks for the geographical areas of interest are set up. Nine references, tabular data, and figures are provided. (Author abstract modified)