We assessed the simple method of measuring red cell deformability described by Reid et al. The technique was found to be reproducible. The validity of the method as a measure of red cell deformation was confirmed by (a) marked reduction of the deformability index after fixation of red cells with glutaraldehyde, and (b) an inverse correlation of deformability index with high-shear blood viscosity (r = 0.4; P <0.001). There was no correlation of deformability index with low-shear blood viscosity, plasma viscosity, fibrinogen, or the white cell count. In normal subjects, deformability index was similar in males and females, and in smokers and non-smokers. Patients with acute myocardial infarction, or intermittent claudication, had reduced deformability compared to controls (P <0.01).