More than half of the patients with angiographically confirmed premature coronary heart disease (CHD) have a familial lipoprotein disorder1. Familial combined hyperlipidaemia (FCHL) represents the most common genetic dyslipidemia with a prevalence of 1.0–2.0% (refs 2,3). FCHL is estimated to cause 10–20% of premature CHD (ref. 1) and is characterized by elevated levels of cholesterol, triglycerides, or both3,4. Attempts to characterize genes predisposing to FCHL have been hampered by its equivocal phenotype definition, unknown mode of inheritance and genetic heterogeneity. In order to minimize genetic heterogeneity, we chose 31 extended FCHL families from the isolated Finnish population5 that fulfilled strictly defined criteria for the phenotype status. We performed linkage analyses with markers from ten chromosomal regions that contain lipid-metabolism candidate genes. One marker, D1S104, adjacent to the apolipoprotein A-ll (APOA2) gene on chromosome 1, revealed a lod score of Z=3.50 assuming a dominant mode of inheritance. Multipoint analysis combining information from D1S104 and the neighbouring marker D1S1677 resulted in a lod score of 5.93. Physical positioning of known genes in the area (APOA2 and three selectin genes) outside the linked region suggests a novel locus for FCHL on 1q21-q23. A second paper in this issue (Castellani et al.) reports the identification of a mouse combined hyperlipidaemia locus in the syntenic region of the mouse genome6, thus further implicating a gene in this region in the aetiology of FCHL.