In early 2024, clade 2.3.4.4b highly pathogenic avian influenza (HPAI) A H5N1 virus was detected in United States dairy cattle. Although so far the public health threat of contemporary clade 2.3.4.4b H5N1 virus strains remains low, continued circulation in mammals and frequent spillover into humans poses a threat of pandemic H5N1. The United States and other countries have stockpiled vaccines and have plans in place to rapidly produce additional vaccine doses should a pandemic H5N1 virus emerge. However, the continued antigenic drift of clade 2.3.4.4b H5N1 antigens compared with historical antigens used by stockpiled vaccines has raised questions of whether these vaccines will confer protection or whether stockpiles need to be updated. We recently evaluated a replicating RNA (repRNA) vaccine against lethal contemporary 2.3.4.4b H5N1 virus challenge in mice and found that a homologous, but not historical, H5 hemagglutinin (HA)–based vaccine conferred protection. Here, we further evaluated the protective capacity of a repRNA expressing the contemporary 2.3.4.4b HA or a repRNA expressing a historical H5 HA (A/Vietnam/1203/2004) in a recently developed lethal nonhuman primate (NHP) challenge model. We found that both vaccines conferred robust protection against lethal 2.3.4.4b H5N1 virus challenge, reducing viral loads and signs of respiratory illness. Our data show that the repRNA platform can elicit protective immunity against lethal influenza virus challenge in NHPs and that historical H5 HAs can elicit cross-protective immunity.