Of 652 consecutive patients referred for coronary angioplasty between September 1980 and March 1984, 49 patients presented with total or functional 'occlusion' of the involved vessel. Total vessel occlusion was defined as absent anterograde filling beyond the lesion. Functional occlusion was defined as faint, late anterograde opacification of the distal segment in the absence of a discernible luminal continuity. In 39 patients, the total or functional occlusion represented a progression, without acute myocardial infarction, of a previously diagnosed stenotic lesion. The maximal potential duration of occlusion was estimated to be 4 weeks or less in 21 patients, more than 4 to 8 weeks in 12, and more than 8 weeks in 16. Dilation of the occluded artery was attempted in the left anterior descending coronary artery in 30 patients, in the right coronary artery in 8, in the circumflex coronary in 7 and in 4 jumpgrafts. For the whole group, angioplasty was successful in 28 patients (57%). The primary success rate with the functionally occluded vessel (81%) was significantly higher than with the total occlusion (45%). In 33 patients with an occlusion estimated to be of 8 weeks or less, angioplasty was successful in 65%. In the 16 patients with an occlusion estimated to be of 8 weeks or less, angioplasty was successful in 65%. In the 16 patients with an occlusion estimated to be of more than 8 weeks duration, dilation was successful in 44%. Of the 21 patients in whom angioplasty was unsuccessful, 11 required surgery (1 urgent with persistent pain and ST elevation and 10 elective). Ten patients were maintained on medical treatment. Of the 28 patients in whom angioplasty was successful, 10 patients had recurrence of symptoms during follow-up (1-42 months). Four were kept on medical therapy, three required bypass surgery and three underwent repeat percutaneous transluminal coronary angioplasty (PTCA). After primary success, late angiographic studies obtained in 20 out of 28 patients showed reocclusion in 8. In conclusion, elective PTCA of totally occluded coronary arteries is feasible but the primary success rate is lower (57%) than that associated with conventional lesions. The long-term clinical results following successful angioplasty are satisfactory (64%), but the incidence of reocclusion is higher (40%).