A spray drift model was developed to deliver real time information to the pesticide applicator. The sprayer is equipped with sensors to deliver real time measurement of operational parameters as spray pressure, boom height, horizontal boom movements and geolocalization. The spray droplet size spectrum as a function of pressure was characterized using PDI measurements. Wind speed and direction were measured using a sprayer mounted 2-D ultrasonic anemometer. For each successive boom position, a diffusion-advection Gaussian tilting plume model is used to compute the spray drift deposits downwind. Drift is computed independently for each droplet classes and each nozzle based on the operating parameters. Field trials were performed on a test plot in various wind conditions. The ground drift was measured for different drift distances using fluorimetry analysis. Results show that drift deposits are mainly affected by wind speed and direction what was correctly accounted for by the model. Short distance drift deposits values were overestimated by the model while long distance drift was underestimated. It appears that this most probably origins from embarked wind speed measurements and diffusion parameter. It is concluded that a treatment of embarked wind speed and diffusion measurement should be used to minimize these errors.