The purpose of this study was to establish a subsampling procedure for benthic macroinvertebrates to aid in the development of a multimetric index to assess the biological condition of streams. Data from six streams that are considered minimally impaired were used. Subsampling was done using a device divided into 24 quadrats. The sediment from each quadrat was sorted, and all organisms were removed and identified. Richness metrics were the most affected by subsample size. Relative-abundance metrics were the most stable, proving that the sample was well distributed throughout the tray and abundance proportions were maintained. The results showed that the macroinvertebrate assemblage present in the six quadrats was similar to that present in the full sample. These analyses indicated that six quadrats, randomly designated, with a minimum of 200 collected specimens could be used in place of the full sample. In routine water management, managing time and costs are a major challenge; therefore, this type of simplification is absolutely necessary to ensure that a biomonitoring tool is useful for practical applications.