Eighty Gram-negative bacilli (54 Enterobacteriaceae and 26 nonfermenting Gram-negative bacilli) obtained from multiple institutions in the United States were distributed in a blinded manner to seven testing laboratories to compare their performance of a test for detection of carbapenemase production, the Carba NP test. The Carba NP test was performed by all laboratories, following the Clinical and Laboratory Standards Institute (CLSI) procedure. Site-versus-site comparisons demonstrated a high level of consistency for the Carba NP assay, with just 3/21 site comparisons yielding a difference in sensitivity (P Ͻ 0.05). Previously described limitations with bla OXA-48-like carbapenemases and bla OXA carbapenemases associated with Acinetobacter baumannii were noted. Based on these data, we demonstrate that the Carba NP test, when implemented with the standardized CLSI methodology, provides reproducible results across multiple sites for detection of carbapenemases.KEYWORDS carbapenemase, antimicrobial resistance, screening, Gram-negative bacilli C linical microbiology laboratories around the globe have instituted a range of rapid screening tests for detection of carbapenemase production in the rising tide of carbapenem-nonsusceptible Gram-negative bacilli (GNB). One method, the Carba NP test, offers a rapid and simple colorimetric method for the detection of carbapenemase activity in isolates of GNB. In principle, the test is similar to the classic -lactamase detection assays described by Escamilla and by Skinner and Wise in the 1970s, in which benzylpenicillin acidimetrically produced a red color after hydrolysis by Haemophilus influenzae -lactamase (1, 2). With the Carba NP test, hydrolysis of imipenem produces acid and drives the pH indicator, phenol red, from red to yellow.The Carba NP test has been studied by several investigators (3-14), including evaluations and comparisons to a commercial version (15)(16)(17)(18)(19)(20). However, the literature lacks consistency in the Carba NP methodology. The test has evolved several times, including alterations to reagents, and has gone from a microtiter plate-based format performed on bacterial protein extracts to a rapid tube-based test with abbreviated processing. Most studies have been single-laboratory evaluations; there is a lack of multicenter performance assessments utilizing a single, harmonized method.As part of an effort to standardize the method, the Clinical and Laboratory Standards Institute (CLSI) Subcommittee on Antimicrobial Susceptibility Testing