Direct torque controlled - DTC induction motor drive involves the separate control of the stator flux and electromagnetic torque through the selection of optimum inverter switching modes. To select directly the optimum switching vectors (switching states), the stator flux and electromagnetic torque errors are used together with the position of the stator flux linkage space vector. In the conventional DTC drive, each switching vector is defined for flux position of p/3 radians (there are only six sectors considered). Due to this, the optimal response can not be obtained, and low-speed operation can be problematic, due to the error in the stator resistance used for stator flux estimation or due to the inadequate switching vector selection table use. To obtain improved performance of DTC drive during start-up or during changes in the reference flux and torque, it was used a fuzzy-logic based switching vector selection process. For this purpose, a Mamdani-type fuzzy logic system was used, and a rule base was formulated. Thus the goal was to use a fuzzy logic system, which improved the system performances, i.e. faster torque and flux responses. For this purpose, during start-up, switching states giving a higher increase in stator flux modulus have to be selected by the fuzzy system, during this time the changes in the torque being small. When the flux error becomes small, switching states which give faster increase in the torque have to be selected by the fuzzy logic system. So, the selection of the appropriate rules base of the fuzzy logic system is presented in this paper. For emphasizes the advantages of the fuzzy logic system over the conventional one, there were presented the experimental results, as well.