arose from Dr. Robison's discussion of the dose assessment work where "average soil concentrations" were used. The Advisory Group believe the concept of averaging should be restricted to doSe estimates and, even in this case, should be used with great care. There are a number of methods that could be used for basing doSe asSesSments on actual soil data: Compute the dose on the basis of highest estimated soil concentration for any unit (of 1/4 or 1/2 hectare size) on the island. This method could result in a very conservative dose estimate if most units on the island had much lower concentrations than the highest unit. Average all the unit (1/4 or 1/2 hectare) average soil concentrations on an island and use this grand average to estimate potential dose. This method weights all units equally, even though some units would undoubtedly have greater utilization by inhabitants than other units. A conServative approach would be to uSe the cleanup criteria designation for the island (40, 100, or 400 pCi/g, for village, agricultural, or picnic islands, respectively). Dose esti- mates for this option were obtained by Robison, et al. in their dose assessment paper presented to us at the meetings. Conservative dose estimates are likely since most units have average soil concentrations much below these cleanup criteria. Perhaps the best method would be to estimate the dose Separately for each 1/4 or 1/2 hectare unit on the island using the estimated average soil concentration for the unit.