Using the real-time diffusion methodology first developed by Codastefano et al. [Rev. Sci. Instrum. 48, 1650 (1977)] we show that deviations from strictly one-dimensional transport (i.e., blockages and voids) widely cited in the literature as leading to erroneous results, have little effect on the measured diffusivity. In this methodology, radiotracer, initially located at one end of a cylindrical diffusion sample, is used as the diffusant. The sample is positioned in a concentric isothermal radiation shield with collimation bores located at defined positions along its axis. The intensity of the radiation emitted through the collimators is measured as a function of time using solid state detectors. Diffusivities are calculated from the signal difference between the detectors. These results were obtained using114mIn radiotracer in benign indium. A 60% blockage was simulated by using a 2 mm source disk diffusing into 3 mm diameter host section. A void/bubble was simulated by inserting a 1 mm diameter by 1 mm long side plug into the sample 7 mm from the top (radiotracer section). The resulting self-diffusivities obtained were, for both modifications, comparable with other ground experiments. These results agree with numerical simulations. An analytical model describing the minimal effects of blockages and voids on the output diffusivity is presented. © 2000 American Institute of Physics.