Patients with B-cell lymphoma may have disease manifestations ranging in size from more than a 1000 cm3 down to the volume of a single cell. If targeted radionuclide therapy is to become a curative treatment, all individual tumor cells must also be eliminated. Given the vast differences in particle energy of different electron- emitting radionuclides, one questions whether the mean absorbed dose is a relevant parameter for use in single-cell dosimetry and whether it would not be more accurate to adopt a stochastic approach to dosimetry. Monte Carlo simulations were performed of energy deposition from 1000, 300, 100, or 10 electrons uniformly distributed in a sphere with a radius of 7.7 µm. The simulated electrons were monoenergetic (18 keV, 28 keV, 141 keV, or 935 keV). The absorbed dose per emitted electron, the absorbed fraction, the fraction of the cellular volume in which energy is deposited, and the dose-volume histograms were calculated. Absorbed fractions varied between 0.60 (18 keV) and 0.001 (935 keV), and the absorbed dose to the cell per electron emitted varied by a factor of 10, from 0.898 mGy (18 keV) to 0.096 mGy (935 keV). The specific energy varied between 0 and 46 mGy for the case showing the best uniformity (1000 18-keV electrons). The nonuniformity of the absorbed dose to a cell increases with increasing electron energy and decreases with the number of decays inside the studied volume. The wide distribution of energy deposition should be taken into account when analyzing and designing trials for targeted radionuclide therapy.