Contradictory results were obtained from previous studies aiming at defining the frequency of Ha-ras codon 12 mutations in bladder tumors. Differences in the sensitivities of the methods used could account for this discrepancy. In this study, we reevaluated the frequency of Ha-ras codon 12 mutations in a series of 87 human bladder tumors using a combination of two different methods. The first was derived from the protocol of Ooi et al and consisted in a one-step allele-specific polymerase chain reaction using mismatched primers in two separate PCR. This method is very rapid and highly sensitive, detecting the presence of minor populations (less than 10%) of mutant alleles. The second strategy consisted in screening all tumors using natural restriction fragment length polymorphism (RFLP) analysis. The two methods were in complete concordance and enabled us to show that only one out of 87 primary bladder carcinomas (1%) exhibited the mutation, in accordance with previous studies. These results strongly suggest that, even if minor cell populations overexpress codon 12 Ha-ras mutation, the analysis of this mutation cannot be used to screen potentially invasive transitional cell tumors of the bladder.