Backgrounds: Anti-Sm antibody is specific to systemic lupus erythematosus (SLE) and is one of the American College of Rheumatology’s criteria for SLE. However, the clinical accuracy varies according to different detection systems. This study is to investigate if a new test kit with synthetic anti-SmD3 peptides has a better diagnostic performance. Methods: We used a fluorescent enzyme immunoassay to determine anti-Sm and anti-SmD levels in a SLE group, a rheumatic disease control (rheumatoid arthritis, RA) and a healthy control (HC) group. Receiver operating characteristic curves were constructed and the cut-off value established was able to increase sensitivity of anti-SmD measurement. Results: In total, 213 SLEs, 75 RAs and 108 HCs were studied. Of these, 5.63% of SLEs, 0% of RAs and 0% of HCs were positive for anti-Sm while anti-SmD was positive in 21.13%, 0% and 3.85% of SLE, RA and HC respectively. The sensitivity of anti-Sm and anti-SmD test kit was 5.63% and 21.13% respectively. With the cut-off value analyzed by receiver operating characteristic (ROC) curve, sensitivity was increased to 27.7%. Conclusions: Anti-SmD antibodies have high specificity for SLE and the test kit using SmD3 antigen has better diagnostic performance. In this study population, we identified 27.7% of SLE patients as being was positive for anti-SmD antibodies.