Unusual elevations of serum creatinine (S-CR) out proportion to increases in serum urea nitrogen (S-UN) are frequently observed in patients with diabetic ketoacidosis when S-CR is measured by the Jaffe end-point reaction. This has been ascribed to interference from acetoacetate but this is not however observed with kinetic DuPont ACA methodology. Eighteen patients with diabetic ketoacidosis were studied: SCR measurements were done using the end point Technicon SMA 6/60 method (Group a, 10 patients) or the kinetic DuPont ACA method (Group B, 8 patients). The values for S-CR in Group A patients (mean value and S-D were 3.3 +/- 1.1 mg/dl) were significantly different from Group B patients (1.6 +/- 0.24 mg/dl) (P less than 0.01). A significant positive correlation was obtained between the "excess anion gap" and S-CR in Group A patients (r = + 0.81, p less than 0.01). The results from two patients in whom serial measurements of S-UN, S-CR and the anion gap were carried out further demonstrate the analytical interference. The study demonstrates that in diabetic ketoacidosis elevated creatinine values measured by an end-point method should not necessarily be interpreted as evidence of significant renal impairment and if possible should be verified by a kinetic method which is free of "ketone" interference.