In order to limit the number of intrusive inspections, the most modern customs administrations rely on risk analysis as the only effective tool for facilitating trade and securing their operations given the important growth in trade volume in recent years. However, customs administrations in developing countries have been slow to adopt this methodology. This paper demonstrates that based on the experience of Senegal, data mining and statistical scoring techniques can be used effectively by customs in developing countries to assess risk and to assign declarations to the various inspection channels. The paper also shows that the in-house development of this type of system by customs administrations in these countries significantly advances the process of modernization.