Nanoparticle-based radio-enhancement has the potential to improve cancer cell eradication by augmenting the photoelectric cross-section of targeted cancer cells relative to the healthy surroundings. Encouraging results have been reported for various nanomaterials, including gold and hafnia. However, the lack of scalable synthesis methods and comparative studies is prohibitive to rationalized material design and hampers translation of this promising cancer management strategy. Here, we present a scalable (>100 g day-1) and sterile alternative to conventional batch synthesis of group IV metal oxides (TiO2, ZrO2, and HfO2), which yields near-monodisperse ultrasmall metal oxide nanoparticles with radio-enhancement properties. Access to group IV oxide nanoparticles, which solely differ in atomic number but otherwise exhibit comparable morphologies, sizes, and surface chemistries, enables the direct comparison of their radio-enhancement properties to rationally guide material selection for optimal radio-enhancement performance. We show that the metal oxide nanoparticles exhibit atomic-number-dependent radio-enhancement in cancer cells (HT1080 and HeLa), which is attenuated to baseline levels in normal fibroblasts (normal human dermal fibroblasts). The observed radio-enhancement effects show excellent agreement with physical dose enhancement and nanoparticle dosimetry calculations. Direct benchmarking against gold nanoparticles, the current gold standard in the field, rationalizes the use of hafnia nanoparticles based on their radio-enhancement performance, which is superior to equi-sized gold nanoparticles. Taken together, the competitive radio-enhancement properties for near-monodisperse nanoparticles produced by scalable and sterile flame spray synthesis offer a route to overcoming key roadblocks in the translation of nanoparticle-based radio-enhancers.