We evaluated phenotypic methods for the detection of kPc carbapenemases in 44 clinical isolates of K. pneumoniae having reduced susceptibility to carbapenems, 30 of which were kPc-positive and 14 kPc-negative. Both the agar dilution and disk diffusion methods were performed for imipenem, meropenem and ertapenem. The following phenotypic methods were assayed: the double disk synergy test, using boronic acid or clavulanic acid as inhibitors, "combined" disks of carbapenem plus inhibitor (boronic acid, clavulanic acid and both boronic plus clavulanic acid), by using a pre-diffusion technique and the modified Hodge test. The double disk diffusion test using boronic acid could detect all kPc-positive isolates, but adjustment of disk distance was necessary for achieving such performance. The simulation of combined disks by our pre-diffusion technique detected all kPcpositive strains for all 3 carbapenems when using boronic acid as inhibitor, clavulanic acid was less susceptible and specific as compared with boronic acid. The modified Hodge test using any carbapenem was clearly positive for all kPc-producing isolates. This test was negative for all kPc-negative strains when imipenem or meropenem were used, but 2/14 isolates yielded a weak positive result when using ertapenem. By measuring the enhanced growth of E. coli aTcc 25922 observed in this test, we could objectively discriminate true-positive (= 8 mm) from false-positive results (< 5 mm). Our results show that the use of phenotypic methods is effective for the rapid detection of kPc producers in K. pneumoniae isolates with reduced susceptibility to carbapenems.