Measurement of neonatal bilirubin using the Jendrassik-Grof method (x) and the EKTACHEM NBIL assay (y) was compared over a 6-month period in a total of 1191 specimens from 483 patients less than 30 days of age. Linear regression analysis of the data yields a slope of 0.937, an intercept of 0.387, and Sy,x of 0.55 and a correlation coefficient of 0.983 for a total of 1032 specimens from patients less than or equal to 14 days of age and a slope of 1.090, an intercept of 0.002, an Sy,x of 1.03 and a correlation coefficient of 0.950 for a total of 159 specimens from patients greater than 14 days of age. The best correlation between EKTACHEM NBIL assay and the reference Jendrassik-Grof method was observed in samples from patients less than or equal to 14 days of age. Data from patients older than 14 days showed a higher proportional bias and a lower correlation coefficient between the methods. High performance liquid chromatographic analysis demonstrated that patients greater than 14 days of age had a higher incidence of elevated delta-bilirubin. Linearity extends to 200 mg/l. The NBIL assay provides a rapid, precise micromethod that is less sensitive than the Jendrassik-Grof method to the in vitro photo degradation of bilirubin and is not subject to the interference from hemoglobin and lipids. Because delta-bilirubin is not measured by this method, it is only recommended for newborns less than or equal to 14 days of age.