The relationship between the 5A/6A stromelysin-1 promoter polymorphism and progression of angiographically determined coronary artery disease (CAD) has been examined in men treated for 32 months with gemfibrozil or placebo in the Lopid Coronary Angiography Trial (LOCAT). The frequency of the 5A allele was 0.40 (95%, CI, 0.36-0.43), and in the sample as a whole 12% of the men were homozygous for the 5A allele. In the placebo group, diffuse progression of disease was, on average, completely prevented in men with the genotype 5A/5A as measured by a 0.30% increase in mean average diameter of the coronary artery segments (ADS), compared with a mean 1.79% decrease in the combined group with the genotype 5A6A or 6A6A (mean +/- S.E.M., +0.007 +/- 0.020 mm vs. -0.043 +/- 0.0.08 mm, P = 0.03). A similar relationship with genotype was seen for disease progression determined by the mean minimal luminal diameter (MLD); with the 5A5A group decreasing by an average of 1.72% compared with 5.54% in the 5A/6A plus 6A/6A group (-0.029 +/- 0.034 mm vs. -0.102 +/- 0.013 mm, P = 0.06). In the gemfibrozil-treated group, the effect on disease progression associated with the 5A/6A alleles was of a similar pattern as in the placebo group, but the effect was less marked and was not statistically significant. This study confirms the previously reported beneficial effect on disease progression associated with the 5A allele and raises the possibility that patients with CAD who are homozygous for the 6A allele, and who represent 25-30% of the population, may be at particular risk of rapid progression of disease and may require particularly aggressive lipid lowering therapy to prevent disease progression.