Restenosis is the main problem limiting long-term success of percutaneous transluminal coronary angioplasty (PTCA) and is most accurately evaluated by follow-up angiography. We compared the primary and long-term results of angioplasty in 268 consecutive patients (293 segments) with first PTCA (PTCA 1, angiographic follow-up 98%) and in 66 patients (76 segments) with repeat PTCA after restenosis (PTCA 2, angiographic follow-up 92%). Forty clinical, angiographic and procedural factors were assessed in relation to outcome. Primary success rate was higher in PTCA 2 (91% vs 67.5%) and major complications were fewer (4.5% vs 16%). Higher inflation pressure (7.9 +/- 2.3 vs 6.8 +/- 1.8 atm, P less than 0.005) and larger balloons (3.5 +/- 0.5 vs 3.2 +/- 0.5 mm, P less than 0.005) were used for PTCA 2, resulting in lesser residual stenosis (33 +/- 16% vs 40 +/- 18%, P less than 0.05). Restenosis rate (greater than or equal to 70%) after PTCA 1 and after PTCA 2 (27% vs 36%, P = NS) and the mean time to recurrence (4.7 vs 5.3 months, P = NS) were similar. Procedural factors were the main determinants of long-term success in primary PTCA. The restenosis risk was independently related to residual stenosis greater than or equal to 45% (P less than 0.001), variant angina (P less than 0.05) and multivessel disease (P less than 0.05) after PTCA 1 and to male sex (P less than 0.001) and higher inflation pressure (P less than 0.05) after PTCA 2. Mild to moderate intimal tearing was associated with less restenosis after PTCA 1, but not after PTCA 2. Including 9 patients (10 segments) with a third PTCA, 70% of the 66 patients with repeat PTCA had a successful long-term outcome. Repeat angioplasty should therefore be considered as an integral part of PTCA therapy. Restenosis however remains a major concern. An optimal primary result with a minimal residual stenosis is decisive for first PTCA, whereas avoidance of a dissection by using lower inflation pressure on a restenosis might improve the long-term outcome of repeat PTCA.