Plant viral diseases are the most important pathogen attack tomato particularly of which is Tomato yellow leaf curl virus (TYLCV). Obtaining stable and durable resistances seems the most effective solution for controlling the disease. In this context, a breeding programme with two steps was conducted in the Niayes ecological area of Senegal: i) 41 tomato varieties from different origins were screened to assess their reaction to TYLCV disease, ii) 12 resistant varieties and two susceptible were tested in an experimental field. The first step was a preliminary trial in unreplicated plots, to identify promising varieties. However, disease assessments (incidence and severity) were done three times and we used longitudinal data analyses to classify the varieties for resistance. Yields in t/ha were also measured for the different varieties. Among our accessions, Roma VF and Xina known for their susceptibility, and Nadira known for its tolerance, were used as controls. Our results revealed 12 resistant varieties, 16 tolerant varieties and 8 susceptible varieties. The yield values were well correlated to the resistance measured by both incidence and by severity right from the second date of symptom observations, at the time of flowering. TYLCV disease is thus one of the main factors limiting tomato-growing the Niayes agro-ecological area. The second step of the selection process was done in a complete block design with three replications. This trial allowed a confirmation of the 12 resistant varieties. The final selection of the varieties took into account the other traits, like yield and traits related to the fruit characters.