We describe a Neural-Geo Temporal Model (NGTM) for urban travel demand generation modelling. NGTM concentrates on the representation of temporal interactions between transportation system, land use patterns, spatial location and population that affect the generation of travel demand in urban areas. NGTM combines Neural Networks (NN) and Geographical Information System (GIS) in order to establish a recursive and non-linear modelling of trip generation. NGTM was applied to a case study in Nagoya City, Japan in order to verify its efficiency. NGTM was capable of predicting the trip generation for future development scenarios, considering the temporal evolution of the urban area.