We evaluated the Abbott Vision system, to assess its suitability for use in the physician's office setting. We compared results from Vision for all the available analytes (alkaline phosphatase, urea nitrogen, cholesterol, creatinine, glucose, uric acid, and triglycerides) with results from the Kodak Ektachem 400 and 700. For all analytes, standard curve linearity, assessed with various dilutions of an analyte-supplemented serum pool, was within the ranges claimed by the manufacturer. Within-run precision (CV) for assays of these analytes ranged from 0.9% to 4.7%, run-to-run precision from 1.1% to 7.3%. Comparisons with other methods were generally very good except for a bias in results for blood urea nitrogen and alkaline phosphatase. Hemolysis (hemoglobin at 2.0 and 3.0 g/L) interferes with results for cholesterol, glucose, triglycerides, and uric acid in serum and whole blood. Bilirubin at 82 mg/L interferes with results for creatinine and triglycerides; at 120 mg/L it interferes with cholesterol, glucose, and uric acid results; and at 170 mg/L it interferes with alkaline phosphatase results. Triglycerides up to 5000 mg/L do not interfere with any of the tests. Calibration of the analyzer was stable for one month. We also compared the performance of a skilled operator with that of an unskilled operator and a physician.