Abstract: TDCs(Tectonically deformed coal) are obviously different from normal coal in physical characteristics. It is one of the key factors needed to be considered in CBM(coalbed methane) reservoir modeling. In order to figure out the issue of TDC thickness prediction, we analyzed the recent progress in TDC identification and thickness prediction using well logs and seismic attribution as inputs. Compared with the interactive recognition using well logs, the method for identifying TDCs based on wavelet multi-scale analysis and cluster analysis has higher accuracy and better reliability. Combining with seismic attributes and machine learning algorithms, one can produce a deterministic prediction of TDC thickness with more accuracy. Combining with seismic attributes and geostatistical stochastic simulation, one can produce a non-deterministic prediction of TDC thickness with higher reliability. Although the prediction of TDC thickness has been proceeded for many years, the prediction of TDC types and spatial characteristics still need further study.