Part I of this article discussed the potential functional importance of genetic mutations and alleles of the human cytochrome P450 2D6 (CYP2D6) gene. The impact of CYP2D6 polymorphisms on the clearance of and response to a series of cardiovascular drugs was addressed. Since CYP2D6 plays a major role in the metabolism of a large number of other drugs, Part II of the article highlights the impact of CYP2D6 polymorphisms on the response to other groups of clinically used drugs. Although clinical studies have observed a gene-dose effect for some tricyclic antidepressants, it is difficult to establish clear relationships of their pharmacokinetics and pharmacodynamic parameters to genetic variations of CYP2D6; therefore, dosage adjustment based on the CYP2D6 phenotype cannot be recommended at present. There is initial evidence for a gene-dose effect on commonly used selective serotonin reuptake inhibitors (SSRIs), but data on the effect of the CYP2D6 genotype/phenotype on the response to SSRIs and their adverse effects are scanty. Therefore, recommendations for dose adjustment of prescribed SSRIs based on the CYP2D6 genotype/phenotype may be premature. A number of clinical studies have indicated that there are significant relationships between the CYP2D6 genotype and steady-state concentrations of perphenazine, zuclopenthixol, risperidone and haloperidol. However, findings on the relationships between the CYP2D6 genotype and parkinsonism or tardive dyskinesia treatment with traditional antipsychotics are conflicting, probably because of small sample size, inclusion of antipsychotics with variable CYP2D6 metabolism, and co-medication. CYP2D6 phenotyping and genotyping appear to be useful in predicting steady-state concentrations of some classical antipsychotic drugs, but their usefulness in predicting clinical effects must be explored. Therapeutic drug monitoring has been strongly recommended for many antipsychotics, including haloperidol, chlorpromazine, fluphenazine, perphenazine, risperidone and thioridazine, which are all metabolized by CYP2D6. It is possible to merge therapeutic drug monitoring and pharmacogenetic testing for CYP2D6 into clinical practice. There is a clear gene-dose effect on the formation of O-demethylated metabolites from multiple opioids, but the clinical significance of this may be minimal, as the analgesic effect is not altered in poor metabolizers (PMs). Genetically caused inactivity of CYP2D6 renders codeine ineffective owing to lack of morphine formation, decreases the efficacy of tramadol owing to reduced formation of the active O-desmethyl-tramadol and reduces the clearance of methadone. Genetically precipitated drug interactions might render a standard opioid dose toxic. Because of the important role of CYP2D6 in tamoxifen metabolism and activation, PMs are likely to exhibit therapeutic failure, and ultrarapid metabolizers (UMs) are likely to experience adverse effects and toxicities. There is a clear gene-concentration effect for the formation of endoxifen and 4-OH-tamoxifen. Tamoxifen-treated cancer patients carrying CYP2D6*4, *5, *10, or *41 associated with significantly decreased formation of antiestrogenic metabolites had significantly more recurrences of breast cancer and shorter relapse-free periods. Many studies have identified the genetic CYP2D6 status as an independent predictor of the outcome of tamoxifen treatment in women with breast cancer, but others have not observed this relationship. Thus, more favourable tamoxifen treatment seems to be feasible through a priori genetic assessment of CYP2D6, and proper dose adjustment may be needed when the CYP2D6 genotype is determined in a patient. Dolasetron, ondansetron and tropisetron, all in part metabolized by CYP2D6, are less effective in UMs than in other patients. Overall, there is a strong gene-concentration relationship only for tropisetron. CYP2D6 genotype screening prior to antiemetic treatment may allow for modification of antiemetic dosing. An alternative is to use a serotonin agent that is metabolized independently of CYP2D6, such as granisetron, which would obviate the need for genotyping and may lead to an improved drug response. To date, the functional impact of most CYP2D6 alleles has not been systematically assessed for most clinically important drugs that are mainly metabolized by CYP2D6, though some initial evidence has been identified for a very limited number of drugs. The majority of reported in vivo pharmacogenetic data on CYP2D6 are from single-dose and steady-state pharmacokinetic studies of a small number of drugs. Pharmacodynamic data on CYP2D6 polymorphisms are scanty for most drug studies. Given that genotype testing for CYP2D6 is not routinely performed in clinical practice and there is uncertainty regarding genotype-phenotype, gene-concentration and gene-dose relationships, further prospective studies on the clinical impact of CYP2D6-dependent metabolism of drugs are warranted in large cohorts.