Even prior to the Coronavirus Disease 2019 (COVID-19) pandemic, the fourth leading cause of death globally was lower respiratory tract infection. Further, upper respiratory infection burden was 17.2 billion cases globally in 2019 alone, reiterating the significant burden of respiratory diseases. Since last year, vaccines for three current major respiratory pathogens are now available: severe acute respiratory coronavirus 2 (SARS-CoV-2), influenza, and respiratory syncytial virus (RSV). Despite the major contributions these vaccines have had in reducing morbidity and mortality, these vaccines are limited in their breadth and durability and leave many susceptible to disease. Efforts are underway for more universal vaccines against these pathogens. Antibody discovery and high-resolution antibody-antigen structures inform vaccine design by characterizing antibodies that are promising for broad and potent protection. Structure-guided vaccine design aims to re-elicit these promising antibodies against complex pathogens by strategically modifying wildtype protein structures. In this dissertation, I utilized structure-based analyses and tools to modify the SARS-CoV-2 spike (S) and influenza hemagglutinin (HA) and silence some epitopes while leaving accessible those that are recognized by antibodies with high potency and breadth. I modified the SARS-CoV-2 S receptor binding domain (RBD) and influenza HA head domain to immune-focus antibodies to select epitopes including antibody epitopes that overlap with receptor binding sites. I designed minimal structures containing promising epitopes where potent antibodies can inhibit binding to the host receptor. I demonstrate the addition of N-linked glycans can influence antibody specificity and improve neutralization in vivo. I have also demonstrated that epitope resurfacing can promote antibody cross-reactivity. The display of our immune-focused antigens by nanoparticle display improves breadth and potency compared to monomers and we demonstrate potency and breadth are highly durable in mice for over a year. The work here supports structure-guided design strategies for the development of more universal vaccines for respiratory viruses and other diverse pathogens.