The observed midday maximum in the mixing ratio of HO2 at Rishiri Island in June 2000 was ~10 pptv, but photochemical box model simulations overpredicted HO2 at this location by an average of 70%. This overestimation was significant only when the mixing ratio of NO was lower than 300 pptv, and was coincident with overprediction of the NO/NO2 ratio. We detected several organoiodines, presumably emitted from seaweeds, and propose the presence of the IO radical. IO could reduce HO2 mixing ratios via the formation of HOI that may subsequently be scavenged by aerosols or lost by photolysis and may also convert NO to NO2 directly. Model calculations with known iodine chemistry could reproduce the observed HO2 with 12--25 pptv of IO. Although iodine chemistry is unlikely to explain the entire discrepancy in HO2, several pptv of IO could significantly reduce HO2 mixing ratios and NO/NO2 ratios.