In combined analysis of two genome-wide association (GWA) studies in cases enriched for family history, we recently identified seven loci (on 1p13.3, 1q41, 2q36.3, 6q25.1, 9p21, 10q11.21 and 15q22.33) that may affect the risk of coronary artery disease (CAD). Apart from the locus on 9p21, the other loci await substantive replication. Furthermore, the effect of these loci on CAD risk in a broader range of individuals remains to be determined. We therefore undertook association analysis of single nucleotide polymorphisms at each locus with CAD risk in 11 550 cases and 11 205 controls from nine European studies. The mean age at first event was 59.5 years (10.0). 79% of cases were men and 59% had had a myocardial infarction.The 9p21.3 locus showed unequivocal association (rs1333049, combined odds ratio (OR) 1.20, 95% CI 1.16 to 1.25; p = 2.81 × 10−21). We also confirmed association signals at 1p13.3 (rs599839, OR 1.13, 1.08 to 1.19; p = 1.44 × 10−7), 1q41 (rs3008621, OR 1.10, 1.04 to 1.17; p = 1.02 × 10−3) and 10q11.21 (rs501120, OR 1.11, 1.05 to 1.18; p = 4.34 × 10−4). The associations with 6q25.1 (rs6922269, p = 0.020) and 2q36.3 (rs2943634, p = 0.032) were borderline. The 15q22.33 locus did not replicate. The 10q11.21 locus showed a sex interaction (p = 0.015), with a highly significant effect in women (OR 1.29, 1.15 to 1.45; p = 1.86 × 10−5) but not in men (OR 1.03, 0.96 to 1.11; p = 0.387). There were no other strong interactions of any of the loci with other traditional risk factors. The loci at 9p21, 1p13.3, 2q36.3 and 10q11.21 acted independently and cumulatively increased CAD risk by 15% (12–18%) per additional risk allele.The findings provide strong evidence for association between at least four genetic loci and CAD risk. Cumulatively, these novel loci have a significant impact on the risk of CAD at least in European populations.