A study was made to determine the uncertainty in subcritical reactivity as inferred from inverse kinetics rod drop experiments (using the three-point method) due to the statistical uncertainty inherent in the observed count rate of the neutron sensor. The two methods employed were a classical propagation of error analysis, and an analysis of simulated repeated rod drops, with an assumption that the uncertainty in reactivity was due to the detection process itself for both techniques, To test the analysis methods, the reactivity uncertainties for various experimental rod drop data sets were computed by both methods. There was excellent agreement of the results. The propagation of error analysis may be used on three-point subcriticality measurements to provide an experimenter with an index to the statistical reliability of the inferred reactivity estimate.