OBJECTIVES: To assess the utility of a new Health Employer Data and Information Set (HEDIS) measure of quality of care of pediatric upper respiratory infections (URI) in an Israeli HMO, develop a comparable measure, and estimate baseline performance. METHODS: Data sources included physician visit and pharmacy purchase files. We selected physician visits in 2003 bearing diagnoses corresponding to ICD-9-CM 460 (URI-nasopharyngitis), 465 (URI other/unspecified site), and 461 (sinusitis) among children ages three-months to 18 years. Visits by children for whom antibiotics had been purchased in the preceding 30 days, or who had additional diagnoses were excluded. Associated antibiotic use was defined as purchase of any antibiotic prescribed by the physician who recorded the URI diagnosis within three days of the visit. RESULTS: A total of 160,781 first encounters for URI or sinusitis met the inclusion criteria. Nine percent were associated with an antibiotic purchase. Antibiotic use for URI or sinusitis increased with age, from 4% of visits in children under the age of one to 14% in children ages 15-18. Sixty-two percent of sinusitis, 7.1% of "nasopharyngitis," and 32% of "unspecified URI" visits were associated with antibiotic use. CONCLUSION: Transfer of quality measures between medical systems requires consideration of differences in medical practice and of the ways in which diagnoses are used and coded. In adapting the HEDIS measures of care of pediatric upper respiratory infections, we chose to examine prescribing practices within subgroups, as well as to explore sinusitis treatment, an issue not currently addressed by HEDIS measures. Overall antibiotic use in conjunction with diagnosis of URI in our population was comparable to reported figures in US commercial and Medicaid plans for the same year, but varied with age and specific diagnosis. Additional attention to sinusitis treatment, not currently addressed by HEDIS, is warranted.