Immune assays are increasingly being used to study immunity in wild animals, with applications in ecology, evolution, and conservation. However, the use of immune assays is hindered by the limited taxonomic breadth of studies that validate assays in non-model organisms and by limited understanding of the variables that can affect assay results. While freezing blood samples in the field for later laboratory analysis is a common practice, the effect of storage conditions on the viability of samples is unclear. In this study, we evaluated the effect of plasma storage conditions (refrigerated or frozen) on the results of two immune assays (bacterial killing assay [BKA] and hemagglutination) in eastern box turtles (Terrapene carolina carolina), a declining species that is threatened by disease. We concurrently tested how individual identity and phenotypic or environmental variables influenced immune assay results. We found that freezing plasma samples for 3-17 days produced more repeatable BKA results than refrigerating the samples for up to 3 days, without significantly affecting average immune performance. However, BKA performance was reduced after holding samples frozen for 3-4 months. Additionally, we found that there was no significant difference in hemagglutination between frozen and refrigerated samples. Furthermore, males in general had higher hemagglutination but lower bacterial killing ability than the females. At an individual level, turtles had repeatable differences in immune activity, and the two immune measures were generally correlated with each other. Our findings indicate that freezing of plasma samples for up to 2 weeks is appropriate for both BKA and hemagglutination immune assays in T. c. carolina, and this may extend to related species. Furthermore, we found that individual and sex differences within a species can affect particular immune assays, and future work should evaluate this in other species.