http://www.uah.es Abstract: We present an automated low cost method for evaluating the visibility of traffic signs. For this propose we define a parameter which evaluates how road signs are seen by drivers at night. Thus, the evaluation is done from inside a vehicle, using the headlamps as light sources and a colour digital camera to capture the signs in sequences acquired as we approach them. The captured frames are then automatically processed with a software which allows us to detect and recognize the signs using Support Vector Machines (SVM) as a novel classification technique. Finally, a parameter for measuring the visibility of signs is obtained from the sequence. As example, this technique has been applied successfully over three different signs with three different degrees of surface deterioration.