Klebsiella pneumoniae is a global threat to healthcare, and despite the availability of new drugs, polymyxins are still an important therapeutic option for this and other resistant gram-negative pathogens. Broth microdilution is the only method that is recommended for polymyxins. In this study, we evaluated the accuracy of a commercial Policimbac® plate in determining the polymyxin B MIC for K. pneumoniae clinical isolates. The results were compared with those of the broth microdilution method according to ISO 16782. The Policimbac® plate had an excellent 98.04% categorical agreement, but unacceptable 31.37% essential agreement rates. Almost 2% of major errors as observed. Additionally, 52.94% of the strains overestimated the MIC at 1 µg/mL. Three isolates were excluded from the analysis due to the drying of the Policimbac® plate. To avoid dryness, we included wet gauze for the test, obtaining a 100% of categorical agreement rate; however, a low essential agreement was maintained (25.49%). In conclusion, the Policimbac® plate was unable to correctly determine the polymyxin B MIC for K. pneumoniae isolates. This low performance may interfere with the clinical use of the drug and, thus, with the result of the patient’s treatment.