The reaction of myeloperoxidase compound I (MPO-I) with chloride ion is widely assumed to produce the bacterial killing agent after phagocytosis. Two values of the rate constant for this important reaction have been published previously: 4.7 x 106 M-1.s-1 measured at 25 degrees C [Marquez, L.A. and Dunford, H.B. (1995) J. Biol. Chem. 270, 30434-30440], and 2.5 x 104 M-1.s-1 at 15 degrees C [Furtmüller, P.G., Burner, U. & Obinger, C. (1998) Biochemistry 37, 17923-17930]. The present paper is the result of a collaboration of the two groups to resolve the discrepancy in the rate constants. It was found that the rate constant for the reaction of compound I, generated from myeloperoxidase (MPO) and excess hydrogen peroxide with chloride, decreased with increasing chloride concentration. The rate constant published in 1995 was measured over a lower chloride concentration range; the 1998 rate constant at a higher range. Therefore the observed conversion of compound I to native enzyme in the presence of hydrogen peroxide and chloride ion cannot be attributed solely to the single elementary reaction MPO-I + Cl- --> MPO + HOCl. The simplest mechanism for the overall reaction which fit the experimental data is the following: MPO+H2O2 ⇄k-1k1 MPO-I+H2O MPO-I+Cl- ⇄k-2k2 MPO-I-Cl- MPO-I-Cl- -->k3 MPO+HOCl where MPO-I-Cl- is a chlorinating intermediate. We can now say that the 1995 rate constant is k2 and the corresponding reaction is rate-controlling at low [Cl-]. At high [Cl-], the reaction with rate constant k3 is rate controlling. The 1998 rate constant for high [Cl-] is a composite rate constant, approximated by k2k3/k-2. Values of k1 and k-1 are known from the literature. Results of this study yielded k2 = 2.2 x 106 M-1.s-1, k-2 = 1.9 x 105 s-1 and k3 = 5.2 x 104 s-1. Essentially identical results were obtained using human myeloperoxidase and beef spleen myeloperoxidase.