The International Society for Clinical Hemorheology organized a workshop to compare three instruments for measuring RBC aggregation: LORCA, Myrenne Aggregometer and RheoScan-A. The Myrenne Aggregometer provides indices at stasis (M) and at low shear (M1), with four indices obtained with the LORCA and RheoScan-A: amplitude (AMP), half-time (T1/2), surface area (SA) above (LORCA) or below (RheoScan-A) the syllectogram, and the ratio (AI) of the area above (LORCA) or below (RheoScan-A) the syllectogram to total area (AI). Intra-assay reproducibility and biological variability were determined; also studied were RBC in diluted plasma and in 1% 500 kDa dextran, and 0.003% glutaradehyde (GA)-treated cells in plasma. All measurements were performed at 37 degrees C. Standardized difference values were used as a measure of power to detect differences. Salient results were: (1) intra-assay variations below 5% except for RheoScan-A AMP and SA; (2) biological variability greatest for T1/2 with other indices similar for the three devices; (3) all instruments detected progressive changes with plasma dilution; (4) the Myrenne and LORCA, but not the RheoScan-A, detected differences for cells in dextran; (5) GA-treatment significantly affected the LORCA (AMP, T1/2, SA, AI), the RheoScan-A (AMP, SA, AI) and the Myrenne M parameter. It is concluded that: (a) the LORCA, Myrenne and the RheoScan-A have acceptable precision and suitable power for detecting reduced aggregation due to plasma dilution; (b) greatly enhanced RBC aggregation may not be sensed by the RheoScan-A while the Myrenne M1 index may be insensitive to minor increases of cell rigidity; (c) future studies should define each instrument's useful range for detecting RBC aggregation.