Comfortable outdoor environment benefits the health of citizens and reduces energy consumption and pollution. This study discusses different outdoor thermal sensation and comfort evaluation methods in severe cold area. The database was from a year-long outdoor thermal comfort survey conducted in Harbin, China. Thermal sensation evaluation was developed using meteorological parameters and three popular thermal comfort indices including Standard Effective Temperature (SET*), Physiologically Equivalent Temperature (PET) and Universal Thermal Climate Index (UTCI). Thermal comfort prediction was developed by the three thermal comfort indices and acceptability. Original thermal sensation scales of SET* and PET were less applicable to predict thermal sensation vote (TSV). Calibrated scales of the three indices were obtained based on linear regression results and probit analysis. The accuracies of calibrated scales of thermal sensation were all below 32.8%. The comfortable thermal sensation range in severe cold area varied from “slightly cool” to “hot”. This calibrated range improved accuracies of thermal comfort predicting by around 20%. The unacceptability appropriate to define comfortable range was 9% on the cold thermal sensation side and 26% on the hot side. Adaptation and local exposure also acted on thermal sensation and comfort apart from factors included in thermal comfort indices. Our results provide practical thermal sensation and thermal comfort scales for severe cold area. The discussions indicate the significance of considering adaptation and local exposure for further improving thermal sensation and comfort predicting.