Human and simian immunodeficiency viruses (HIV and SIV), influenza virus, and hepatitis C virus (HCV) have heavily glycosylated, highly variable surface proteins. Here we explore N-linked glycosylation site (sequon) variation at the population level in these viruses, using a new Web-based program developed to facilitate the sequon tracking and to define patterns (www.hiv.lanl.gov). This tool allowed rapid visualization of the two distinctive patterns of sequon variation found in HIV-1, HIV-2, and SIV CPZ. The first pattern (fixed) describes readily aligned sites that are either simply present or absent. These sites tend to be occupied by high-mannose glycans. The second pattern (shifting) refers to sites embedded in regions of extreme local length variation and is characterized by shifts in terms of the relative position and local density of sequons; these sites tend to be populated by complex carbohydrates. HIV, with its extreme variation in number and precise location of sequons, does not have a net increase in the number of sites over time at the population level. Primate lentiviral lineages have host species-dependent levels of sequon shifting, with HIV-1 in humans the most extreme. HCV E1 and E2 proteins, despite evolving extremely rapidly through point mutation, show limited sequon variation, although two shifting sites were identified. Human influenza A hemagglutinin H3 HA1 is accumulating sequons over time, but this trend is not evident in any other avian or human influenza A serotypes.