The aims of this study were to examine the relationships between CYP2D6 genotype and metoprolol dose, S-and R-metoprolol concentrations and clinical effects in patients with systolic heart failure. Data were obtained for 52 subjects, of which 27 had 2 functional alleles (24/27, CYP2D6*1/*1), 22 had 1 functional allele (18/22, CYP2D6*1/*4) and 3 had no functional alleles (CYP2D6*4/*4). Median dose-adjusted concentrations of S-metoprolol (active) were 6.3-and 3.2-fold higher in subjects with zero or one functional allele (P ¼ 0.016 and P ¼ 0.006), respectively, compared with subjects with two functional alleles. For the R-enantiomer (inactive), these concentrations were 10.7-and 3.7-fold higher (P ¼ 0.013 and P ¼ 0.003), respectively. Despite clear gene-concentration differences, no relationships between CYP2D6 genotype and dose or clinical effects could be shown. Although the number with no functional alleles was too small (n ¼ 3) to show effects, in patients with 1 functional allele other sources of variance are likely to be obscuring differences in clinical effects.