Abstract: Objective To evaluate the performance of five routine rotavirus detection assays-gold Immunochromatography assay (GICA), enzyme-linked immuno sorbent assay (ELISA), reverse transcriptase polymerase chain reaction (RT-PCR), real-time RT PCR and polyacrylamide gel electrophoresis (PAGE). Methods First, one rotavirus group A positive sample was 10-fold serially diluted from 1101 fold to 1108 fold and tested by above mentioned assays for lower limit of detection (LLD) evaluation. Second, nine samples, which were confirmed to be rotavirus negative, of poliovirus vaccine strain type Ⅰ to Ⅲ, enterovirus 71, coxsackie virus A16, ECHO virus 6, norovirus type Ⅰ, norovirus type Ⅱ and astrovirus were used to evaluate the specificities of the five assays. Third, a total of 184 infantile diarrhea stool samples were collected and tested with the five assays for sensitivity and specificity evaluation. Results The LLDs of the five assays were different from each other by five orders of magnitude, i.e. 10-6 dilution for real-time RT PCR,10-3 dilution for GICA,10-2 dilution for ELISA and PAGE and 10-1 dilution for RT-PCR. The five assays had no cross reactions to the negative control samples. Based on the PAGE detection results of the clinical samples, the sensitivity was 95.45% for real-time RT-PCR,93.94% for GICA,96.97% for ELISA and 62.12% for RT-PCR, and the specificity was 82.20% for real-time RT-PCR,86.44% for GICA,82.20% for ELISA, 94.07% for RT-PCR. Conclusion PAGE, the classic assay for rotavirus detection, had the highest specificity and would be the best choice for the laboratory confirmation. Real time RT-PCR showed higher sensitivity and specificity than the other assays and could be used for the high throughout detection of rotavirus group A. GICA and ELISA assay are good screening assays because of their low cost. RT-PCR was proved to be with high specificity but low sensitivity and need improvement.