Eulerian models are widely used in air quality modeling. Such models are based on a three dimensional spatial grid in which emissions are released. Thus emissions are instantaneously diluted within the grid cells and the near-source impacts of large point and line sources cannot be properly resolved. Therefore, in regional scale simulations, the approximation made by the model on the emissions from point and line sources can have a significant impact, in particular with large grid cells. Plume-in-grid (PinG) models combine an Eulerian model and a plume or puff model. This coupling offers a better representation of sources within in grid cells, because it uses the plume/puff model to compute locally the dispersion of pollutants from the sources. In a standard PinG model, Gaussian puffs (or plumes) are released with a certain frequency to model with a time discretization the evolution of a plume. When a predefined criteria is reached, based on the number of time steps, the size of the puff or the ratio of the puff concentration to the background concentration, pollutants are transferred to the Eulerian model, which can then compute the dispersion at a regional scale. The main purpose is to be able to use Gaussian local precision at a regional scale. Plume-in-grid model, already exist for point sources (e.g., AMSTERDAM, Polyphemus). They have been used in many studies and have been proven to be efficient. Modeling roadway traffic with point sources would required to spatialy discretized road section with point sources, which would induce a significant increase of the computational burden. Although point sources are convenient to model dispersion from a chimney, line sources are better suited for roadway traffic pollution modeling because it allow having the Gaussian precision with only one source per road section. A new plume-in-grid model that uses Gaussian line source models imbedded within the Eulerian model has been developed. Emissions are treated as steady-state plumes that are released from the line sources. A challenging part was to adapt a steady-state model, which by definition assumes no change with time, to a time dependant Eulerian model. The transfer of pollutants from the Gaussian model to the Eulerian model differs significantly from that of a puff model. It occurs in several cells by distributing the pollutant to be transfered into all cells that should be affected by the plume. The model was developed as a part of the modeling platform Polyphemus. It has been fully tested with some k-tests and results of several validation case studies will be presented.