24The aim of this study was to determine the performance of both cefotaxime and ceftazidime 25 containing agars on the specificity and sensitivity for chromosomal plasmid AmpC harboring Escherichia coli compared to ESBL-producing E. coli and E. coli without 27 ESBL, pAmpC or cAmpC hyperproduction. Second, we evaluated the influence of adding 28 cefoxitin to these agars for detection of both chromosomal AmpC-hyperproducing and plasmid 29 AmpC harboring E. coli. 30 Four different homemade screening agars with cefotaxime (1mg/L), ceftazidime (1mg/L), 31 cefotaxime (1mg/L) with cefoxitin (8mg/L), and ceftazidime (1mg/L) with cefoxitin (8mg/L) were 32 compared to each other for the identification of AmpC producing E. coli. A total of 40 isolates 33 with plasmid encoded AmpC β-lactamases, 40 isolates with alterations in the 34 promoter/attenuator region of the AmpC gene leading to hyperproduction of the β-lactamase, 35 40 isolates with ESBL genes and 39 isolates lacking both a AmpC and ESBL genotype were used 36to test the four agars. 37 The sensitivity and specificity were 100% (95% confidence interval (95% CI) 96.1% to 100%) and 38 48.1% (95% CI 38.6%-60.2%), respectively, for the cefotaxime agar; 100% (95% CI 96.1% to 39 100%) and 49.41% (95% CI 39.8%-61.4%), respectively, for the ceftazidime agar; 96.3% (95% CI 40 89.1% to 99.2%) and 77.2% (95% CI 66.7%-85.2%) respectively, for the cefotaxime with cefoxitin 41 agar; 98.8% (95% CI) 92.6% to 99.6%) and 81.0% (95% CI 70.9%-88.3%) respectively, for the 42 ceftazidime agar with cefoxitin. The main reason for false-positive results were ESBL-harboring 43 strains that grew on various agars; therefore, the specificity of each agar reported here was 44 3 influenced mainly by the proportion of ESBL isolates tested. In conclusion addition of cefoxitin 45 to cefotaxime and ceftazidime containing agars had little influence on sensitivity, but increased 46 specificity for the detection of AmpC in E. coli. 47 48 59 Enterobacteriales (AmpC-E).60