Aim: To evaluate the accuracy rates of acute otitis media (AOM) diagnosis in the community by analyzing the initial AOM diagnosis (according to information from the referral letters to the Pediatric Emergency Department PED ) among children with AOM diagnosis confirmed by tympanocentesis at PED. Methods: A descriptive, retrospective study conducted during the year 2000, including 590 children aged 0-48 mo diagnosed with AOM at PED. AOM was confirmed by tympanocentesis. We defined: (1) simple AOM as < 3 or < 4 episodes during the last 6 or 12 mo, respectively; (2) complicated AOM as ≥ 3 or ≥ 4 episodes during the last 6 or 12 mo, respectively, and/or not improving with antibiotics. Results: 571 (96.8% ) were < 24 mo of age; 355 (60% ) were Arab Bedouins and 235 (40% ) Jews. Seventy-three pediatricians, 57 general practitioners and 27 family practitioners provided referral letters. Information on disease severity was available in 471/590 (79.8% ) patients, of whom 320 (68% ) had simple and 151 (32% ) complicated AOM. Three hundred and sixty-five (62% ) of 590 patients were accurately diagnosed by referring physicians, more frequently in patients with complicated than simple AOM (p lt; 0.001). Diagnostic accuracy was similar among physician specialties (p=0.33) and ethnic groups (p=0.11). Multivariable logistic regression analysis showed that complicated AOM, adjusted for age, sex, ethnic origin, and physician specialty, was the only predictive factor for accurate AOM diagnosis (odds ratio 4.0, p<0.001). Conclusions: Inaccurate diagnosis of AOM in community settings is common, occurring more frequently in simple than in complicated AOM, and is unrelated to physician specialty and children’ s ethnicity. Aim: To evaluate the accuracy rates of acute otitis media (AOM) diagnosis in the community by analyzing the initial AOM diagnosis (according to information from the referral letters to the Pediatric Emergency Department PED) among children with AOM diagnosis by tympanocentesis at PED. Methods: A descriptive, retrospective study conducted during the year 2000, including 590 children aged 0-48 mo diagnosed with AOM at PED. AOM was confirmed by tympanocentesis. We defined: (1) simple AOM as <3 or <4 episodes during the last 6 or 12 mo, respectively; (2) complicated AOM as ≥ 3 or ≥ 4 episodes during the last 6 or 12 months, respectively, and / or not or improving with antibiotics. Results: 571 (96.8%) were <24 mo of Seventy-three pediatricians, 57 general practitioners and 27 family practitioners provided referral letters. Information on disease severity was available in 471/590 (79.8%) patients, of whom 320 (68%) had simp Three hundred and sixty-five (62%) of 590 patients were accurately diagnosed by referring physicians, more frequently in patients with complicated than simple AOM (p lt; 0.001). Diagnostic accuracy was similar Among the physician specialties (p = 0.33) and ethnic groups (p = 0.11). Multivariable logistic regression analysis showed that complicated AOM, adjusted for age, sex, ethnic origin, and physician specialty, was the only predictive factor for accurate AOM diagnosis (odds ratio 4.0, p <0.001 Conclusions: Inaccurate diagnosis of AOM in community settings is common, occurring more frequently in simple than in complicated AOM, and is unrelated to physician specialty and children’s ethnicity.