We evaluated the Technicon DPA-1 immunoassay analyzer on its analytical characteristics. Therefore we studied three assays: albumin in cerebrospinal fluid and IgG and transferrin in serum. When tested with the Cusum test for linearity albumin, IgG, and transferrin measurements showed no deviation from linearity. Closer examination revealed an abrupt difference of recovery (from 99 to 87%) in the albumin assay when the automatic dilution changed over from the primary analytical to the high analytical concentration range. One calibration was found sufficient for at least 14 days of measurement. Imprecision was well below the critical limits for reproducibility. We found reasonable agreement between the results from t h e DPA-1 and the results from comparison methods. However, the correlation plot of IgG showed lack of fit at a distinct segment of the regression line. This appeared to be caused by the poor recovery of the DPA-1 at the lower limit of the high analytical concentration range. The assays of IgG and transferrin were found insensitive for interference by hemoglobin, triglycerides, urea, and bilirubin. The albumin assay was found sensitive for bilirubin and triglycerides. No reagentand sample-to-sample carry-over could be detected in the assays evaluated.