THIS STUDY aims to evaluate the impact of crop rotations on weed dynamics under no-till conditions. Four crop rotations were tested:Wheat/Lentil/Wheat (W/L/W), Lentil/Wheat/Wheat (L/W/W), Triticale/forage Pea/Wheat (T/P/W) and forage Pea/Triticale/Wheat (P/T/W), during the 2016/17, 2017/18 and 2018/19 crop years. The experiment was conducted on the experimental site belonging to the ITGC-FEA of Setif (Algeria). The results indicate that the weed flora is composed of 13 species belonging to 09 families. Dicotyledons are mostly observed during the three campaigns, with proportions of 77.77%, 66.66%, 84.62%, respectively for the campaigns 2016/17, 2017/18 and 2018/19. Annual grasses (monocotyledons) consisted of Avena sterilis and Bromus sterilis. The Shannon and Simpson indices show controversial values depending on the treatments applied. The W/L/W crop rotation was distinguished by both specific diversity and low dominance, conversely the P/T/W treatment, shows high dominance accompanied by low diversity. The linear model relating to the evolution of monocots as a function of precipitation shows that they are negatively and significantly correlated with precipitation in January. The latter is as follows: Y= -0.251x + 34.461 with (R²= 0.998).