Endocrine-disrupting chemicals are exogenous substances that can impact the reproduction of fish, potentially by altering circulating concentrations of 17β-estradiol (E2), testosterone (T), and 11-ketotestosterone (11-KT). Common methods to measure steroids in plasma samples include radioimmunoassays (RIAs) and enzyme-linked immunosorbant assays (ELISAs). The present study examines variability in E2, T, and 11-KT across 8 laboratories measuring reference and pulp mill effluent-exposed white sucker (Catostomus commersoni) plasma. We examine the contribution of assay type (RIA vs ELISA), standardized hormone extraction, location of values on the standard curve (upper and lower limits), and other variables on the ability to distinguish hormone levels between reference and exposed fish and the impact of these variables on quantitation of hormones in different laboratories. Of the 8 participating laboratories, 7 of 8 and 7 of 7 identified differences between sites for female E2 and female T, respectively, and 7 of 7 and 4 of 5 identified no differences between male T and male 11-KT. Notably, however, the ng/mL concentration of steroids measured across laboratories varied by factors of 10-, 6-, 14-, and 10-fold, respectively. Within laboratory intra-assay variability was generally acceptable and below 15%. Factors contributing to interlaboratory variability included calculation errors, assay type, and methodology. Based on the interlaboratory variability detected, we provide guidelines and recommendations to improve the accuracy and precision of steroid measurements in fish ecotoxicology studies.