Climate change is one of the greatest environmental hazards to mankind. The emission of greenhouse gases has resulted in a continuous increase in the temperature of the atmosphere leading to Global warming. CO2 continues to be the leading contributor to the greenhouse effect, with transport being a major CO2 emission source. The majority of transport emissions are from road transport i.e. vehicular emissions. To control vehicular emission, first, an efficient emission monitoring system is required. Direct sensor installation in individual vehicles is neither cost-effective nor the data is easy to collect. In this paper, a scalable vehicle CO2 emission prediction model is proposed which uses vehicle On-Board Diagnostics (OBD-II) port data. The proposed system uses real-time in-vehicle sensor data to estimate CO2 emission of the vehicle using a Recurrent neural network (RNN) based Long short-term memory(LSTM) model. OBD-II dongles can be used to easily transmit the vehicle’s sensor data to the cloud, where the LSTM model uses this data to estimate the real-time CO2 emission of the vehicle. The proposed model provides a scalable and efficient system to monitor emissions at a vehicular level and, has been evaluated using public OBD-II dataset. The details of the data collection, sensors, and adapters used along with vehicle information are outlined in Section V-A .