To investigate the relationship between left ventricular cavity geometry and dynamic intracavitary left ventricular obstruction occurring during dobutamine stress echocardiography, we evaluated 44 patients who underwent dobutamine stress echocardiography. Dynamic intracavitary left ventricular obstruction appeared in 10 (23%) patients (Group 1), and was absent in the remaining 34 patients (Group 2), during dobutamine stress echocardiography. The ratio of the apicalto-narrowest mid-left ventricular minor-axis dimension (D1/D2), as an index of the left ventricular cavitary geometry, was measured in the baseline two dimensional apical long-axis view. The left ventricular out-flow blood Doppler velocity was measured in the apical long-axis or five-chamber view at each stage during dobutamine infusion. Left ventricular end-systolic dimension was smaller (21 +/- 4 mm v 29 +/- 6 mm; p< .01), percent fractional shortening of the left ventricle was greater (51 +/- 4% v 39 +/- 8%, P>.01), and D1/D2 was greater (1.13 +/- 0.08 v.0.96 +/- 0.09; p < .0001) in Group 1 than in Group 2 in the baseline echocardiogram. Maximal out flow blood Doppler velocity during dobutamine infusion correlated moderately with percent fractional shortening of the left ventricle (r = .53; P < .01) and D1/D2 (r = .60; P < .0001). Dynamic intracavitary left ventricular obstruction, defined as an elevated intracavitary Doppler flow velocity with late-systolic peaking, was not related to the presence of left ventricular hypertrophy or the development of hypotension during dobutamine infusion. We conclude that dynamic left ventricular intracavitary obstruction during dobutamine infusion is associated with a characteristic (narrower-waisted) left ventricular cavitary geometry and increased contractility.