Conflicting results are to be found in the literature on the relationship between the M235T polymorphism of the angiotensinogen (AGT) gene and hypertension. The controversy may be due to insufficient numbers of subjects, the variability of the inclusion criteria and the different genotype analysis methods used. We have experienced that the most frequently used, original polymerase chain reaction (PCR)-restriction fragment length polymorphism (RFLP) method involves significant uncertainties when the TT genotype is determined, independently of the restriction digestion. To make the determination more accurate, we improved the PCR by designing a new antisense primer containing only one mismatch instead of the two in the original protocol and also by adding DMSO to the PCR reaction mixture. The original and our improved methods were compared by using DNA from 123 patients: parallel determinations resulted in values of 33 MM, 90 MT and 0 TT with the original method and of 33 MM, 56 MT and 34 TT with the improved RFLP protocol. In summary, a plausible explanation for some of the conflicting data published on AGT M235T polymorphism may be that inaccuracies arose during the determination of the genotype.