Objectives: Annexin A3 (ANXA3) is a potential marker for prostate cancer (PCa). We aimed to develop robust immunoassays suitable for quantifying ANXA3 in urine samples obtained following digital rectal examination (DRE) in order to facilitate the diagnostic performance evaluation of this marker. Design and methods: Anti-ANXA3 monoclonal antibodies were generated and their epitopes mapped. Two different ANXA3 assay prototypes were established on the VIDAS (R) automated immunoanalyser and analytical validation was carried out using post-DRE urine samples obtained from patients with PCa (n = 23) or benign prostate hyperplasia (n = 31). Results: The assays had the same capture antibody (TGC44) but different detection antibodies (13A12 or 5C5), recognizing novel distinct epitopes. Both had a lower limit of quantification <1 ng/mL and were highly specific for ANXA3, not cross-reacting with other annexins. Interassay imprecision was <= 11% and <= 15% for 13A12 and 5C5 assays, respectively. Surprisingly, a total lack of correlation was observed between ANXA3 levels measured by these two assays in post-DRE urines, indicating detection of distinct antigenic variants. Two freeze-thaw cycles did not affect analyte stability in either assay, whereas a lack of stability of antigenic variants was observed when samples were stored at -80 degrees C for 1 month. Conclusions: Two different antigenic variants of ANXA3 are present in post-DRE urines and their clinical significance for diagnosis of prostate cancer should be further investigated. These variants are not stable over time in samples preserved at -80 degrees C. Until this issue is resolved, ANXA3 should only be measured in freshly collected samples.