To facilitate the disposal of municipal solid wastes (MSWs) via thermo-chemical approaches, the accurate measurement of MSWs’ higher heating value (HHV) plays a key role. This study aims to forecast the HHV of MSWs using the optimized multi-variate grey model (OBGM (1, N)) due to its high accuracy under the condition of scanty data. Results show that POBGM (1, 5) with proximate analysis data is the most accurate model with the least error of 5.41% MAPE (mean absolute percentage error). This model also illustrates that ash is the most important factor affecting HHV due to its significant fraction in MSWs, followed by volatiles, fixed carbon and water contents, respectively. With prediction interval (PI) method, most of the actual data can be included by using the 95% confidence intervals.