Objective To explore the influence of prozone effect on anti-nuclear antibodies (ANA) testing by indirect immunofluorescence assay (IIFA).Methods The samples with high titer of ANA (≥1∶1 000) were selected from 880fresh serum samples, and were subsequently diluted in 1∶100, 1∶1 000and 1∶10 000ratio.Prozone effect was defined as fluorescence intensity from 1∶1 000dilution was stronger than that from1∶100dilution.The samples with prozone effect were determined manually or by Sprinter XL and EUROPattern.The samples with prozone effect were further characterized by combinations of fluorescence patterns, fluorescence intensities and autoantibody specificities.Results A total of 880samples were tested.Importantly, 34samples displayed prozone effect (3.86％in total and 29.57％in samples with ANA≥1∶1 000).Interestingly, prozone effect was identified by manual detection as well as by Sprinter XL with similar fluorescence patterns and fluorescence intensities.Notably, EUROPattern can only select the central area for identification.Among all samples with prozone effect, 74.42％samples exhibited fluorescence intensities of≥1∶10 000.Speckled pattern was the most prevalent fluorescence patterns in samples with prozone effect (46.51％).In addition, anti-RNP antibodies (62.79％) were the most popular autoantibodies in samples with prozone effect, followed by anti-dsDNA antibodies (51.16％) and anti-SSA antibodies (51.16％).Conclusion Prozone effect was present in ANA testing, especially in samples with high titers, resulting in underestimating the titers.The study highlighted that special attention should be paid to the prozone effect in clinical practice.