BACKGROUND: Adherence to previous national cholesterol guidelines has been low. We assessed whether lipid screening and management was consistent with the National Cholesterol Education's ATPIII in a sample of primary care practices participating in a quality improvement study. METHODS: Demographic and clinical data were abstracted from charts of 5071 patients aged 21 to 84 years, which were seen between June 1, 2001, and May 31, 2003, at 60 practices. Clinical sites were non-university-based primary care practices from 22 North Carolina counties. A dyslipidemia screening test was defined as a lipid profile performed on persons when not on a lipid-lowering drug. Among patients receiving a lipid profile, the proportion of patients appropriately treated, per ATPIII, was calculated. Practice level variation in screening and management was examined using the 50th (20th and 80th) percentile values across practices. RESULTS: The median practice level dyslipidemia screening rate during the 2 years was 40.1% (25.8%, 53.7%) of their age-eligible patients. The appropriate decision regarding lipid-lowering therapy was documented within 4 months of the lipid profile for 79.3% (69.0%, 86.0%) of practices' patients. Within 4 months, among the drug-ineligible patients, 100% (94%, 100%) were not prescribed drugs; 33.3% (6.3%, 50.0%) of the drug-eligible patients were prescribed lipid-lowering agents. CONCLUSIONS: The median dyslipidemia screening rate met the recommendations. There remains a need to improve the management of dyslipidemia; in particular, there was a significant underprescription of lipid-lowering drugs.