The relation between left ventricular end-systolic wall stress (ESWS) and the velocity of circumferential fibre shortening (Vcfs) was examined in four non-random groups with 12 patients each. In group A, preload was increased by the administration of hydro- xyethylstarch 30 mL min−1. In group B, preload was reduced by administering nitroglycerine 4.0 ± 0.8 mg h−1. A change in pulmonary capillary wedge pressure by 3 mmHg was taken to be the end point for preload manipulation. To assess a change in the relation between ESWS and Vcfs, patients in group C received adrenaline at an infusion rate of 4 g min−1. Patients in group D served as the controls. Geometric variables of ESWS and Vcfs were determined by transoesophageal echocardiography. A linear model was used to assess the relation between ESWS and Vcfs within each group by regression analysis, and analysis of covariance performed to detect significance of intragroup and intergroup differences (P < 0.005). No significant changes were found during preload intervention. With adrenaline, Vcfs increased significantly for a given ESWS. It is concluded that the relation between ESWS and Vcfs, in a multiple patient setting, is independent of modest preload changes and may have the potential to indicate inotropic effects.