Draw resistance is felt by the consumer when his cigarette is lit, not before. In spite of this obvious fact, this physical parameter is usually measured on unlit cigarette. In order to assess more accurately the smoker's feeling, the theoretical aspects of the draw resistance measurement during smoking have been studied and an experimental device has been developed. During a puff, the pressure drop across the cigarette is influenced by several parameters which are independent on the characteristics of the cigarette itself. Theory and experiments show that the dead volume and the clogging up of the Cambridge filter, in the sample holder, significantly modify the measured pressure at the mouth end. The minimization of these effects can be reached by the calculation of a single parameter, more specifically characteristic of the draw resistance of the cigarette, and derived from the evolution of pressure versus time. In addition, another way for the draw resistance assessment can be proposed. This way consists in the integration of the measured pressure profile during each puff, and gives a value related to the necessary energy to supply for the generation of each of them. Using a mono-canal smoking machine, a cigarette holder with a connection for a pressure sensor, an A/N converter and a computer, an experimental device has been developed. For the calibration of the system, a procedure using pressure drop standards composed of capillary tubs has been proposed. Finally, the developed device enables the measurement and the recording of the draw resistance during smoking, and then a better assessment of the smoker's feeling.