In this work, we studied the effect of bismuth oxide particle size and its attenuation capacity as a filler additive in epoxy resins. Six samples were prepared according to the amount of microparticles and nanoparticles in the sample and were coded as ERB-1, ERB-2, ERB-3, ERB-4, ERB-5, and ERB-6. One of the composite epoxies contained Bi 2 O 3 microparticles at a 50:50 ratio (ERB-6) and was chosen as the control composite, and the number of microparticles (MPs) was gradually decreased and replaced by nanoparticles (NPs) to produce epoxy-containing Bi 2 O 3 nanoparticles at a 50:50 ratio (ERB-1). The morphological and thermal characteristics of the studied composites were tested. The attenuation capability of the prepared composites, which is determined by the Bi 2 O 3 particle size, was determined experimentally using a semiconductor detector, an HPGe-detector, and three different gamma-ray point sources (Am-241, Co-60, and Cs-137). The linear attenuation coefficient (LAC) of ERB-3, which contained 30% nanoparticles and 20% microparticles, had the highest value compared to the other composites at all the energies discussed, while the ERB-6 composite had the lowest value at all energies. The radiation-shielding efficiency (RSE) of the prepared samples was determined at all discussed energies; at 662 keV, the radiation-shielding efficiency values were 15.97%, 13.94%, and 12.55% for ERB-3, ERB-1, and ERB-6, respectively. The statistics also proved that the attenuation capacities of the samples containing a combination of nanoparticles and microparticles were much superior to those of the samples containing only microparticles or nanoparticles. A ranking of the samples based on their attenuation capacity is as follows: ERB-3 > ERB-4 > ERB-2 > ERB-1 > ERB-5 > ERB-6.