Crayfish are widely recognized as an important ecological component of stream systems, but there has been limited work to develop and evaluate the reliability of sampling methods for lotic crayfishes, especially efforts that are temporally and spatially comprehensive. We desired a quantitative method to assess crayfish communities in streams with rocky substrate. Our objective was to develop a method to obtain and compare representative density estimates with acceptable variance and reasonable effort, and to illustrate use of the method by using it to 1) describe and compare diurnal habitat associations by lotic crayfish, and 2) detect density changes over time. Our study encompassed four sites on two rivers, two seasons, and 8 y (1991–1998) to evaluate a 1-m² quadrat sampler, and a sampling method that stratified effort among five macrohabitats to reduce variability. This method performed well for both seasons, detecting spatial differences among macrohabitats and temporal differences among years. Spatial differences were expressed as macrohabitat selectivity by the crayfish community, and showed a consistent trend across streams and seasons. In particular, macrohabitats with slower current velocities consistently had the highest densities. Temporal differences included documentation of decreased densities in several macrohabitats across 5 y. Sampling precision, measured by coefficients of variation, was acceptable but not considered high. Statistical power was good for detecting spatial differences, but reduced and variable for detecting temporal changes. Our findings 1) demonstrate the use of stratifying quantitative sampling for lotic crayfish communities by habitats, 2) confirm the importance of evaluating sampling methods, and 3) illustrate the consistent way in which Ozarks crayfish communities used available macrohabitats.