This in vitro study examined the combined effects of double antibacterial drugs against multidrug-resistant Pseudomonas aeruginosa (MDRP). The tested clinical isolates from Hiroshima University Hospital were 40 strains which met the criteria for MDRP, that is, the minimum inhibitory concentration (MIC) was > or = 16 microg/mL of meropenem, > or = 4 microg/mL of ciprofloxacin and > or = 32 microg/mL of amikacin. Using the original checkerboard plates for colistin (CL), arbekacin (ABK), aztreonam (AZT), rifampicin (RFP) and piperacillin (PIPC), MIC values were determined for single and double combinations. Based on the MIC values, fractional inhibitory concentration index values were calculated and the combined effects (synergy action or additive action) were evaluated. The three strongest drugs among the tested combinations were i) CL + RFP (synergy, 80.0%; additive, 17.5%), ii) RFP + ABK (synergy, 7.5%; additive, 70.0%) and iii) RFP + AZT (synergy, 5.0%; additive, 77.5%). In these cases, the arithmetic mean MIC value of each drug significantly decreased as follows: i) 1.38 microg/mL (alone) and 0.26 microg/mL (with RFP) for CL, 19.85 microg/mL (alone) and 1.85 microg/mL (with CL) for RFP; ii) 19.85 microg/mL (alone) and 7.53 microg/mL (with ABK) for RFP, 8.87 microg/mL (alone) and 2.79 microg/mL (with RFP) for ABK; iii) 19.85 microg/mL (alone) and 10.15 microg/mL (with AZT) for RFP, 28.3 microg/ mL (alone) and 6.65 microg/mL (with RFP) for AZT. Of 40 strains, metallo-beta-lactamase and aminoglycoside 6'-N-acetyltransferase were found in 20 and 37 strains, respectively; however, no significant influence of these factors was observed on the combined effects of i), ii) and iii). The results of this study provide an in vitro rationale for RFP plus CL, ABK or AZT as an effective combination therapy for MDRP infections, although the results should be verified and compared with other antibacterial drugs in further studies.