Diversity indices are commonly used to measure changes in marine benthic communities. Although diversity indices give a good understanding of how biological communities respond to changes in sediment, responses of these indices can vary because of small sample sizes, and insensitive responses can result from an inappropriate choice of test community. This study focused on species density and two indices of community structure and explored how they respond to variations in sediment characteristics in our target region, Japan, and in two local areas within this region. Our analysis of the Japanese regional dataset showed a decrease in species density and the dominance of a few species as sediment conditions degraded. Local case studies showed that species density responded to sediment degradation at sites where the community was variable. However, the indices for community structure sometimes became misleading because of inaccurate estimates with small sample sizes, and could become insensitive because of masking by community variability. We conclude that species density is a sensitive indicator of change for marine benthic communities, but that indices of community structure should only be used when there is a sufficient sample size and a test community is distinguishable from other coexisting communities.