We used intravascular ultrasound (IVUS) to assess the accuracy of manufacturers' stent balloon compliance charts. Many interventional cardiologists rely on manufacturers' compliance charts to select stent size and optimize stent diameters according to inflation pressures during percutaneous procedures. We randomly selected 212 patients who had de novo coronary lesions that had been treated with a single, bare metal, > or =3.0-mm stent (Bx velocity, NIR, TETRA/PENTA, S660/S670/S7) under IVUS guidance. Cases of stent overlap and postdilatation with another balloon were excluded. Predicted stent diameters were derived from each manufacturer's compliance charts, and stent size and final maximal deployment pressures were derived from each physician's report. IVUS-measured minimum stent diameters (range 1.4 to 4.0 mm, mean 2.79 +/- 0.48) were smaller than predicted diameters (range 3.1 to 4.57 mm, mean 3.79 +/- 0.44). The ratio of IVUS to predicted diameters ranged from 44% to 97% (mean 74 +/- 10%). This finding was common to all 3 stent sizes: 74 +/- 12% for 3.0 mm, 73 +/- 9% for 3.5 mm, and 74 +/- 9% for 4.0-mm stents (p = 0.9). This finding was also common to all 4 stent manufacturers, 72 +/- 8% for Boston Scientific, 76 +/- 11% for Guidant, 73 +/- 9% for Cordis, and 74 +/- 11% for Medtronic (p = 0.13), and to different stent lengths. Only 3.8% of the stents achieved 90% of the predicted minimum stent diameters, and only 24.6% achieved 80% of the predicted minimum stent diameters. In conclusion, in human coronary arteries, minimal stent diameter measured by IVUS is significantly smaller than that predicted by in vitro compliance charts. These differences are independent of stent manufacturer, length, diameter, and deployment pressure.