OBJECTIVE--To compare the usefulness of magnetic resonance imaging (MRI) and Doppler ultrasound with that of cross sectional echocardiography and oscillometric blood pressure measurement for the evaluation of aortic coarctation after surgical repair. DESIGN--Prospective study. Aortic diameters measured by cross sectional echocardiography, MRI, and angiography (selected cases) and functional data determined by physical examination, oscillometric blood pressure measurement, and continuous wave Doppler. SETTING--Tertiary referral centre. PATIENTS--40 patients aged 2-28 years (mean 10.6 years) who had had surgical correction of aortic coarctation (mean follow up 5.7 years). RESULTS--In all patients MRI gave diameter measurements of the aortic arch and the thoracic aorta whereas in half of them cross sectional echocardiographic measurement of the isthmic region failed. The correlation coefficient for aortic diameters measured by MRI and angiography was 0.97 and that between MRI and echocardiography was 0.89. Peak velocities in the descending aorta correlated better with residual narrowing of the aortic isthmus or distal aortic arch or both than systolic blood pressure gradients between the upper and lower limbs. A peak velocity of < 2 m/s in the descending aorta during systole excluded important restenosis. Prolongation of anterograde blood flow during diastole always indicated a morphological abnormality--either important restenosis or aneurysmal dilatation. CONCLUSIONS--MRI was better than cross sectional echocardiography for imaging the aortic arch after coarctation repair and measuring its diameter. Peak velocity in the descending aorta correlated better with residual stenosis than did the systolic blood pressure gradient between the upper and lower limbs and this index could be used to indicate a need for MRI.