Abstract. A gap in the leptospirosis field remains the lack of well-characterized sample collections that allow for comparison of new methods to standard ones. In the context of a flood-related outbreak of leptospirosis evaluated in Anuradhapura, Sri Lanka, a specimen bank was obtained with detailed metadata accompanied by gold standard diagnostic test results. Blood samples collected on admission and 14 days later from suspected cases of leptospirosis were tested using microscopic agglutination test (MAT) (17 serovars), an in-house enzyme-linked immunosorbent assay (ELISA) using a locally obtained strain of Leptospira kirschneri as sonicated antigen, a commercially available ELISA based on sonicated Leptospira biflexa, and a quantitative polymerase chain reaction (qPCR) assay targeting the pathogenic Leptospira-specific 16S rRNA gene. Of 62 patients presenting within the first 2 days of illness, 31 had confirmed leptospirosis based either on paired-sample MAT or qPCR. During the acute phase, qPCR was most sensitive, detecting 74% of definitively diagnosed cases; immunoglobulin G (IgG) ELISA (in-house), IgG ELISA (commercial), and MAT had sensitivities of 35.5%, 12.0%, and 22.6%, respectively, in detecting definitively diagnosed cases using acute phase serum. Of 40 patients with paired sera, 10 were qPCR positive. Of these, five samples were negative by paired-sample MAT. Of the 11 MAT-positive samples, only five were detected using qPCR confirming that both tests are needed for maximal sensitivity. Regional leptospiral serovar-specific IgG ELISA was superior to MAT. Knowing the regionally dominant serovars improves serological sensitivity in the analysis of acute specimens by ELISA, but qPCR was most sensitive in this patient population.