Background and aims <p>Studies of litter decomposition using standard substrates are important to improve our understanding of how soils function in different ecosystems.</p> Methods <p>In this study, the Tea Bag Index (TBI) method was validated in a subtropical city in southern China. Time series decomposition data were obtained and then used to investigate the performance of the TBI method in soils from a peri-urban forest site and two urban green space sites.</p> Results <p>The results showed that the mass loss curve of green tea leveled off from 30 to 45&#xa0;days. The stabilization factor (S) after 45&#xa0;days was generally not significantly different between retrieval sets, indicating that the incubation period could be significantly reduced in the study area. The S values in the peri-urban forest soil were significantly lower than those in the soils from the two urban green spaces. However, the decomposition rates (<i>k</i>_TBI) in the peri-urban forest soil were also the lowest in the three soils, indicating that the green tea and rooibos tea did not decompose as predicted by the TBI method. The standard TBI method may underestimate the decomposition rate of rooibos tea compared to the results derived from the asymptotic regression.</p> Conclusion <p>It is proposed that the TBI method be modified into the professional version (TBI-pro), which must include time series data of decomposition, and the public version (TBI-pub), which can use the single time point design to maintain its simplicity and practicality, but must be calibrated by the professional version.</p>