Results of an experimental investigation indicate an improvement in accuracy of radial bucklings derived from activation distributions measured in reflected cylindrical systems can be obtained if: resultant activities are fitted to radial spatial functions derived from homogeneous two-group diffusion theory (i.e., Activity (R) = A J0(λR) + C I0(βR), where λ2 = radial buckling), and activation distributions are measured with a detector whose ratio of is high. Radial bucklings derived from activation distributions measured with In, Au and Cu foils in the same core showed that values derived from the In data were the least sensitive to the region of the analyzed. On the basis of a two-group model, radial activation distributions measured with a detector in a reflected core which satisfies the following conditions , where S1 = fast-thermal coupling coefficient, will yield a J0 distribution only, because the increase in activity from the increase in thermal flux is cancelled by the decrease in activity from the decrease in fast flux near the core-reflector boundary. Conclusions are substantiated by theoretical predictions based on the radial variation of fluxes calculated from two-group homogeneous diffusion theory.