In order to determine the relationship of restenosis following percutaneous transluminal coronary angioplasty (PTCA) to risk factors such as hypercholesterolemia, hyperglycemia, smoking, and weight, we performed a univariate analysis to test the association of these variables with restenosis in 723 patients who had percutaneous transluminal coronary angioplasty and follow-up catheterization. Cholesterol levels were higher in younger and female subjects (less than 0.0001). Initial cholesterol did not predict restenosis, and follow-up cholesterol levels showed an inverse relationship with restenosis (P less than .02). There was a trend (P less than .09) toward decreased restenosis in those who were active smokers at the time of follow-up catheterization. No differences were seen in diabetics with hyperglycemia, in both treated and untreated groups (P = NS). A stepwise multiple logistic regression was used to simultaneously test the association of the above risk factor variables to restenosis. None of the interactions were found to be significant, except cholesterol at follow-up (P = .001). Therefore, the status of serum cholesterol, blood sugar, smoking, and weight during the time of PTCA and at follow-up catheterization may be unimportant in predicting restenosis. Thus, we conclude 1) that to better determine the effect of these variables on restenosis, they should be estimated at times other than follow-up and 2) that the pathophysiological mechanism of restenosis may have different risk factors than progression of atherosclerotic coronary artery disease.