Freshwater aquaculture system is one of the main sources of nitrous oxide (N2O) and must be accounted for when reporting indirect N2O emissions from agricultural industries. The IPCC based approach recommend the indirect N2O emission factors (EF5) to estimate the emissions from freshwater. Unfortunately, the EF5 for lake aquaculture, one of the most common freshwater aquacultures, has never been reported due to the scarcity of field data. To better understand the magnitude of EF5 at lake aquaculture and identify the control factors, the EF5 at aquaculture farm and open water (non-aquaculture region) of Lake Taihu were investigated based on long-term (2012–2017) in-situ field measurements. Our results showed the indirect N2O emission at the aquaculture farm (1.52 ± 0.49 μmol m-2 d-1) was over one order of magnitude higher than at the open water (0.12 ± 0.49 μmol m-2 d-1). Furthermore, we also found large variability in the EF5, which varied by one order of magnitude across time. The EF5 was predicted by nitrogen and the mass ratio of carbon to nitrogen. The significantly higher mass ratio of DOC to DIN resulting from feed application contributed to substantial increase in the N2O production efficiency and EF5 at the aquaculture farm. The average EF5 at the aquaculture farm and open water were 0.0021 ± 0.0013 and 0.0013 ± 0.0010, respectively. However, the measured EF5 was lower than that IPCC’s default value of 0.0025, implying IPCC method yielded the overestimated indirect emission, and large bias will occur when only use constant value considering the dramatic variability of observed EF5.