Row crops are a kind of typical vegetation canopy between discrete canopy and continuous canopy. Kimes et al. studied the directional thermal radiation of row crops using the geometrical optical model, which simplified row structure as “box” and neglected the gap among foliage and did not consider the emissivity effects. In this work we take account of the gaps along illumination and viewing directions and propose a bi-direction gap model on the basis of the idea of gap probability of discrete vegetation canopy introduced by “Li-Strahler” and inter-correlation of continuous vegetation developed by Kuusk. It can be used to explain “hot spot” effects in thermal infrared region. The gap model has been validated by field experiment on winter wheat planted in shape of rows and results show that the gap model is better than Kimes’ model in describing the directionality of thermal infrared emission for row crops. Row crops are a kind of typical vegetation canopy between discrete canopy and continuous canopy. Kimes et al. Studied the directional thermal radiation of row crops using the geometrical optical model, which simplified row structure as “box ” and neglected the gap among foliage and did not consider the emissivity effects. In this work we take account of the gaps along illumination and viewing directions and propose a bi-direction gap model on the basis of the idea of ​​gap probability of discrete vegetation canopy introduced by “Li-Strahler ”and inter-correlation of continuous vegetation developed by Kuusk. It can be used to explain“ hot spot ”effects in thermal infrared region. The gap model has been validated by field experiment on winter wheat planted in shape of rows and results show that the gap model is better than Kimes’ model in describing the directionality of thermal infrared emission for row crops.