Patients over a wide anion gap range demonstrated no significant chloride bias from coulometric titrations. Bromide at 0.7 mmol/l gave a 1 mmol/l chloride response (4). Selectivity of chloride cartridge (mmol/l of anion required to increase chloride results by 1 mmol/l) (4). A list of substances and conditions known to affect the level of sodium, potassium or chloride in vivo is given by both Young et al (5) and Friedman et al (6). No representation is made by Roche Diagnostics regarding the completeness of these lists or the accuracy of the information contained therein. References: (1) Glick MR, Ryder KW, Glick SJ. Interferographs:Users Guide to Interferences in Clinical Chemistry Instruments, 2nd ed. Indianapolis, IN: Science Enterprises Inc; 1991:35-44. (2) Tietz NW. Fundamentals of Clinical Chemistry. Philadelphia, Pa: 3rd ed. WB Saunders Co; 1987:617. (3) Tietz NW. Clinical Chemistry. Philadelphia, Pa: 2nd ed.,WB Saunders Co; 1994:1356, 1364, 1365. (4) Simmons M, Sugahara K, Watanabe M. Development of an Improved Chloride Electrode for the Boehringer Mannheim/Hitachi 700 Series Analyzers. Poster session presented at the 1988 National Meeting of the American Association for Clinical Chemistry, New Orleans, LA, July 27, 1988 (reprints available from Roche Diagnostics Corporation). (5) Young DS, et al. Clin Chem. 1975;21, No 5. (6) Friedman RB, et al. Clin Chem. 1980;26, No 4.