A DD genotype of the angiotensin I-converting enzyme gene has been implicated in various diseases. However, genotype frequencies differ between previous reports, and data on the association of DD genotype with disease are sometimes conflicting. Although elimination of mistyping is of crucial importance, assessment of the accuracy of currently adopted typing methods has rarely been performed. Mistyping of the DD genotype is reported to occur by a conventional method with insertion/ deletion (I/D) flanking primers using polymerase chain reaction (PCR). We investigated whether currently adopted genotyping methods by PCR are reliable or not. We genotyped 248 patients by conventional PCR methods with I/D flanking primers with or without dimethyl sulfoxide (DMSO), and confirmed the DD genotype with insertion-specific primers with or without DMSO. Mistyping occurred frequently, not only in both methods without DMSO but also in a modified method with I/D flanking primers with inclusion of DMSO. Typing by these methods proved to lead to erroneous results more frequently than had been previously thought. To reduce mistyping frequency, initial PCR genotyping with I/D flanking primers with an inclusion of DMSO, followed by confirmation of the DD genotype by insertion-specific primers with DMSO, is recommended.