Hydroperoxy radical (HO2) measurements made during the second Southern Ocean Atmospheric Photochemistry Experiment (SOAPEX-2) were used as a modeling case study to investigate the role of aerosol uptake of free radicals in remote marine boundary layer (MBL) air. The SOAPEX-2 campaign was held at the Cape Grim Baseline Atmospheric Pollution Station on the northwestern tip of Tasmania during austral summer from 18 January to 18 February 1999. A box model based on the Master Chemical Mechanism (MCMv3) was tailored to campaign conditions. With no aerosol uptake of HO2 the model overestimated the midday (1100--1400 hours) HO2 measurements by 74--83% on two clean marine air days with HO2 measurements. Two different methods were used to simulate aerosol uptake of free radicals. The first method used a simple first-order rate coefficient for interfacial mass transport (kinetic regime), kkin, with two treatments of the aerosol surface area, estimated from clean MBL Californian air and calculated by analysis of lognormal aerosol number versus size distributions from the first Aerosol Characterization Experiment (ACE-1). The second method used a rate coefficient which takes into account diffusion to the surface as well as interfacial mass transport (transition regime), ktrans, integrated over most of the aerosol size range (up to 2.5 ¿m diameter) using aerosol number size distribution data from ACE-1. With the simple uptake rate the model to measured overestimation of HO2 was on average 63 and 20% with uptake coefficients of 0.2 and 1, respectively. With ktrans the overestimation was 60 and 44%, respectively. These overestimations are upper limits because the calculations did not take into account the largest sea salt mode (which was not measured). An estimate of the aerosol in the largest sea salt modes was calculated and used to recalculate the values of ktrans, reducing the overestimation to 58 and 25%, respectively. The variation in the available literature values for the HO2 uptake coefficient is a large source of uncertainty in the calculated rate of uptake. Another source of uncertainty exists in the values assumed for the aerosol volume in this work. As these data did not exist for SOAPEX-2, we calculated ktrans from ACE-1 aerosol volume data and used a constant averaged value of ktrans; in reality there will be a significant degree of temporal variation in this parameter. Given these uncertainties, we conclude that aerosol uptake of the hydroperoxy radical (HO2) could be a significant process in clean MBL environments and its incorrect parameterization or absence in atmospheric models could contribute to overestimation of measured concentrations.