The heterogeneous reactivity of chlorine peroxide (ClOOCl) on halide-doped ice surfaces was studied using discharge-flow mass spectrometry over the temperature range 209-258 K. By monitoring Cl2O2+ parent ion signals, reactions were observed on ice doped with I-, Br-, Cl-, or F- ions. XCl was the only gas-phase product detected in each reaction, where X = dopant halide. Other products evidently remain bound at the ice surface and were not detected. The heterogeneous reaction ClOOCl + X- --> XCl + products is proposed. Chloride-doped ice layers were made either by freezing aqueous solutions of HCl or NaCl on the flow tube walls or by adding HCl gas to pure ice layers. Thermodynamic considerations indicate that reaction surfaces below the NaX/H2O systems' eutectic points were mixtures of two solid phases (ice and NaX hydrate crystals), while liquid layers existed at HCl/ice surfaces under our conditions. Our measured ClOOCl reaction probabilities on solid surfaces ranged from gamma = 0.026 to gamma greater than or equal to 0.3 on NaI/ice surfaces, from gamma = 0.0016 to gamma = 0.064 on NaBr/ice surfaces (possibly temperature dependent), from gamma = 0.0009 to gamma = 0.0092 on NaCl/ice surfaces, and from gamma = 0.0006 to gamma = 0.0050 on NaF/ice surfaces. ClOOCl reaction probabilities measured on HCl/H2O liquid layers were as high as gamma = 0.0065, and on pure ice surfaces as high as gamma = 0.0011. In the absence of halide ions, there was no detectable reaction of ClOOCl on ice doped with nitric or sulfuric acid (gamma less than or equal to 4 x 10(-4)). Chlorine peroxide uptake was heavily time-dependent, especially on HCl/ice surfaces, indicating that these measurements of heterogeneous kinetics were limited by saturation and deactivation effects in the flow tube. Stratospheric reaction probabilities may therefore differ from those reported here.