Experiments presented in this contribution demonstrate a heterogeneous source of several chlorine oxides, in particular OClO and ClClO2, from ClO radicals passed over water-ice surfaces at low pressures. ClO radicals were generated in a flow system and reaction products were monitored by UV-vis absorption spectroscopy and time-of-flight mass spectrometry using electron-impact and resonance-enhanced multiphoton ionization, respectively. In all experiments an efficient release of OClO into the gas phase was observed upon ice evaporation. No such gas-phase products directly formed in the initial interaction of ClO with the ice surface were detected. To explain these findings, it is proposed that a ClO.H2O complex is formed in a rapidly established equilibrium between ClO monomers and gas-phase H2O. The existence of this complex as well as a surface-enhanced ClO-recombination process is assumed to be responsible for the observed efficient reactive uptake of ClO radicals onto the ice surface. Several possible reaction pathways resulting in the formation of the experimentally observed products are presented. As an alternative pathway, H2O-facilitated gas phase disproportionation of ClO yielding hypochlorous and chlorous acid and subsequent deposition on the surface is considered. The observation of OClO evolving from an ice surface previously exposed to ClO radicals, as well as the lack of any symmetric ClOOCl dimer formation in the presence of high water mixing ratios, carries some possible atmospheric implications: First, there is currently a missing source of OClO in the chemically perturbed polar vortex. The ClO + BrO reaction is presently believed to be the only source of OClO in the stratosphere, although several studies show this reaction system to severely underestimate OClO production in this atmospheric subsystem. It is suggested that this experimentally observed heterogeneous source of OClO could carry implications for the total atmospheric OClO budget. Second, the ClO.H2O complex could directly or indirectly affect the ClOOCl formation rate and thus strongly impact homogeneous chlorine chemistry.