Thermal stress increases community diversity, community variability, and the abundance of potentially pathogenic microbial taxa in the coral microbiome. Nutrient pollution, such as excess nitrogen can also interact with thermal stress to exacerbate host fitness degradation. However, it is unclear how different forms of nitrogen (nitrate vs. ammonium/urea) interact with bleaching-level temperature stress to drive changes in coral microbiomes, especially on reefs with histories of resilience. We used a 13-month field experiment spanning a thermal stress event in the Austral summer of 2016 on the oligotrophic fore reef of Moorea, French Polynesia to test how different forms of nitrogen (nitrate vs. urea) impact the resistance and resilience of coral microbiomes. For Acropora, Pocillopora, and Porites corals, we found no significant differences in diversity metrics between control, nitrate-, and urea-treated corals during thermal stress. In fact, thermal stress may have overwhelmed any effects of nitrogen. Although all three coral hosts were dominated by the bacterial clade Endozoicomonas which is a proposed beneficial coral symbiont, each host differed through time in patterns of community diversity and variability. These differences between hosts may reflect different strategies for restructuring or maintaining microbiome composition to cope with environmental stress. Contrary to our expectation, post-stress microbiomes did not return to prestress community composition, but rather were less diverse and increasingly dominated by Endozoicomonas. The dominance of Endozoicomonas in microbiomes 10 months after peak sea surface temperatures may suggest its ability to utilize host metabolic products of thermal stress for a sustained competitive advantage against other microbial members. If Endozoicomonas is a beneficial coral symbiont, its proliferation after warm summer months could provide evidence of its ability to mitigate coral holobiont dysbiosis to thermal stress and of resilience in coral microbiomes.