Sir, The emergence of Enterobacteriaceae that can produce AmpC b-lactamases is a major worldwide concern. They have arisen through acquisition of a plasmid-mediated ampC gene or hyperproduction of an inducible chromosomal AmpC enzyme, either by induction or derepression. These organisms are resistant to penicillins, first-, secondand third-generation cephalosporins and b-lactamase inhibitor/b-lactam combinations. Therapeutic options to treat these infections are severely limited and often not included in empirical therapy guidelines for serious infections. Rapid detection of AmpC-producing Enterobacteriaceae is therefore essential in order that patients can be optimally managed at an early stage without indiscriminate use of broad-spectrum antibiotics. Methods available for the detection of AmpC producers are relatively unsatisfactory and there are no established guidelines, as there are for the detection of extended-spectrum b-lactamase (ESBL)-producing organisms. We evaluated the commercial D69C AmpC Detection Disc Set (Mast Group Ltd, UK). This system is based on a combination disc method that incorporates cefpodoxime (CPD10) as the screening agent in the presence of AmpC-inducing or -inhibiting agents (exact formulations undisclosed), enabling the detection of both plasmid-mediated and chromosomal AmpC, whether inducible or derepressed. An ESBL inhibitor is additionally included to unmask and minimize interference from any ESBL that may be present on the AmpC detection system. Three discs are used: A (CPD10+AmpC inducer), B (CPD10+AmpC inducer+ESBL inhibitor) and C (CPD10+AmpC inducer+ESBL inhibitor+AmpC inhibitors). A total of 103 previously characterized isolates (comprising 80 AmpC+, 15 AmpC2/ESBL+, 1 AmpC2/KPC+ and 7 AmpC2/ ESBL2 strains) were blindly tested using the detection kit. These isolates were a combination of reference strains and clinical isolates that had been characterized by genotypic or disc diffusion methods, and included strains with chromosomal derepressed, inducible and plasmid-mediated AmpC resistance mechanisms. The Enterobacteriaceae represented were Escherichia coli, Klebsiella pneumoniae, Morganella morganii, Citrobacter freundii, Enterobacter cloacae, Enterobacter aerogenes and Serratia marcescens. Testing and interpretation were performed according to the manufacturer’s instructions. Isolates were sub-cultured onto Mueller–Hinton agar to achieve semi-confluent growth. The three discs were placed on the plate as far apart as possible, and the plate was incubated at 378C in air for 18–24 h. The organisms were interpreted as ‘AmpC negative’ if all zone sizes differed by ≤3 mm and as ‘AmpC positive’ if the zone size of disc C exceeded those of both discs A and B by at least 5 mm. If zone sizes of both discs B and C exceeded that of disc A by at least 5 mm and the zone sizes of discs B and C had a difference of ,5 mm, the isolate was classified as ‘AmpC negative, but exhibits a different resistance mechanism’. As shown in Table 1, The D69C kit correctly determined the AmpC status of all 80 AmpC+ and 23 AmpC2 isolates tested. Amongst the 23 AmpC2 isolates, 12 out of the 15 AmpC2/ ESBL+ isolates and the 1 AmpC2/KPC+ isolate were correctly classified as ‘AmpC negative, but exhibits a different resistance mechanism’. All seven AmpC2/ESBL2 isolates were correctly identified as ‘AmpC negative’. The kit failed to detect the presence of a different resistance mechanism in 3 out of 15 AmpC2/ESBL+ isolates—they were an SHV ESBL+ K. pneumoniae, a TEM/CTX-M ESBL+ E. coli and a TEM/SHV+ ESBL+ E. coli. There is no immediate explanation for these results, particularly since several other TEM, SHV and CTX-M ESBL+ isolates were correctly identified as ‘exhibiting another resistance mechanism’. The ESBL activity might have been masked by the AmpC inducer (incorporated into all three discs), and hence these results highlight an important limitation of the D69C kit, i.e. it is solely a method for AmpC detection, and its use cannot be extrapolated to ESBL detection. The ESBL inhibi-