During epidemics, such as the frequent and devastating Ebola virus outbreaks that have historically plagued regions of Africa, serological surveillance efforts are critical for viral containment and the development of effective antiviral therapeutics. Antibody serology can also be used retrospectively for population-level surveillance to provide a more complete estimate of total infections. Ebola surveillance efforts rely on enzyme-linked immunosorbent assays (ELISAs), which restrict testing to laboratories and are not adaptable for use in resource-limited settings. In this manuscript, we describe a paper-based immunoassay capable of detecting anti-Ebola IgG using Ebola virus envelope glycoprotein ectodomain (GP) as the affinity reagent. We evaluated seven monoclonal antibodies (mAbs) against GP-KZ52, 13C6, 4G7, 2G4, c6D8, 13F6, and 4F3-to elucidate the impact of binding affinity and binding epitope on assay performance and, ultimately, result interpretation. We used biolayer interferometry to characterize the binding of each antibody to GP before assessing their performance in our paper-based device. Binding affinity (KD) and on rate (kon) were major factors influencing the sensitivity of the paper-based immunoassay. mAbs with the best KD (3-25 nM) exhibited the lowest limits of detection (ca. mug mL(-1)), while mAbs with KD > 25 nM were undetectable in our device. Additionally, and most surprisingly, we determined that observed signals in paper devices were directly proportional to kon. These results highlight the importance of ensuring that the quality of recognition reagents is sufficient to support desired assay performance and suggest that the strength of an individual's immune response can impact the interpretation of assay results.