The inter-individual variation of response to cancer chemotherapy and radiation therapy has a substantial genetic component and is a subject of pharmacogenetic studies. Our goal is to develop assays to analyze selected single nucleotide polymorphisms (SNPs) in genes with proven relevance for pharmacogenetics. Today's research in the field needs technological platforms for which the demands both in terms of quality and throughput are high. Pharmacogenomics utilizes various genotyping techniques as well as gene-expression studies in studying the effect of different pharmaceuticals as well as the pathways that influence them. Here we provide an overview of the platforms currently in use, and discuss their efficiency, precision and technological stability. The amount of hands-on time needed and the cost-effectiveness are discussed. We also compared the success rate of several methods. SNaPshot is a homogeneous-phase-based primer extension method and SNPstream UHT is a solid-phase-based primer extension method. The Nanogen chip employs binding of a biotinylated probe to a chip coated with hydrogel enriched with streptavidin. This universal design presents the possibility to perform both genotyping as well as gene-expression experiments, even on the same chip. Results were compared with the classical enzymatic restriction enzyme fragment polymorphism (RFLP) assay. The initial set included eight SNPs in the X-ray cross-complementing group 1 protein, XRCC1; APEX nuclease (multifunctional DNA repair enzyme) 1, APEX1; multidrug resistance gene, MDR1; gluthation-S-transferase P1, GSTP1; and tumor protein p53, TP53. Functional SNPs in these genes have been shown to influence the therapeutic effect of a series of anti-cancer drugs such as 5-fluorouracyl, doxorubicin, etoposide, thioguanine and irinotekan. Three series of 36 breast cancer patients treated with 5-fluorouracyl and mitomycin, of 92 breast cancer patients treated with doxorubicin and of 109 control individuals have been genotyped. This resulted in 1527 genotypes. Of these, 391 had been analysed by SNaPshot, 900 by SNPstream UHT, 454 by Nanogen and 966 by RFLP. The success rate was 91% for SNaPshot, 97% for SNPstream UHT, 96% for Nanogen, and 99% for RFLP. When comparing the methods pairwise, the lowest variation was observed between SNPstream UHT and Nanogen (0.4%). SNPstream UHT and Nanogen had similar rates of discrepancy when compared with RFLP: 3.2% and 2.6%, respectively. Close re-analysis of those samples revealed false calls due to incomplete cutting during the RFLP. The highest observed dis-concordance observed was between SNaPshot and SNPstream UHT (8.9%). Careful evaluation of genotyping quality is needed both for large epidemiological studies and smaller clinical trials, and especially in prospective pharmacogenetic studies of separate individuals in clinical settings.