Environmental DNA (eDNA) has become an important tool for monitoring and studying aquatic biodiversity. However, the consistency of results across different laboratories remains insufficiently evaluated. In this study, three independent laboratories processed and sequenced the same batch of freshwater samples under a unified and standardized framework, with positive controls used to validate workflow reliability. The results showed that no false negatives were detected across laboratories in positive controls, and false-positive sequences accounted for less than 0.5%. For environmental samples, the species detection overlap among laboratories was ≥ 80%, with a coefficient of variation in α-diversity of less than 0.05. The relative abundance of commonly detected species exhibited strong correlations across laboratories (r ≥ 0.94). Although significant differences in community composition were observed (PERMANOVA, R2 = 0.35, p = 0.005), the abundance patterns of dominant species were highly consistent. This study demonstrates that, under standardized key procedures, eDNA metabarcoding can achieve consistent species detection and diversity estimation across laboratories, while highlighting that comparisons of community structure should be interpreted cautiously due to potential laboratory-specific effects.