Résumé : One of the most important skills in our lives is to learn to adapt to the different changes that occur in our surroundings, finding the best strategy to keep obtaining our goals and reaching our aims. This is known as cognitive flexibility, and it is the ability to rapidly change our behavior according to the changes in our environment. It is a very important cognitive function which shows impairments in a wide range of neurodevelopmental and psychiatric conditions, and neurodegenerative diseases that imply affections in common areas of the brain, where one of these affected common areas is the striatum.The striatum is the main input nucleus of the basal ganglia, which is a group of sub-cortical nuclei involved in motor control, learning, attention, and cognitive flexibility. The two efferent subpopulations of striatal projection neurons (SPNs) have been studied comprehensively regarding their involvement in motor control, and several models have been proposed trying to explain the different interplay between the direct (d-) and the indirect (i-) SPNs pathways in locomotion and action selection. However, different studies have demonstrated their specific role in cognitive flexibility when assessing their behavioral outputs after d- and i-SPNs activity alterations or ablation in flexible contexts, and the different models proposed may remain too simplistic to explain d- and i-SPNs interplay during changes in contingencies. In addition, SPNs also have a role in regulating and encoding the execution of sequences of actions, and how this regulation and encoding changes in a flexibility context has not been described yet. The goal of this study was to characterize throughout days of training the evolution of single d- and i-SPNs activity in the dorsal striatum in two different operant flexibility challenges: Fixed-ratio 1 to 5 (FR1-5) and a reversal learning exercise (REVL), and assess how do the different activity changes correspond to the claimed d- and i-SPNs specific roles in flexibility concluded from a collection of studies where their activity was modified in different ways. In general, we observed a higher and more stable activity for d-SPNs than for i-SPNs, where d-SPNs appeared modulated in similar ways across the flexibility exercises while i-SPNs showed a more variable modulation. In addition, both SPNs populations showed specific activity patterns emerging for each training of FR1-5 and REVL. Furthermore, we found differences in the activity and encoding of SPNs between FR1-5 and REVL. The results obtained during this study allowed us to set a hypothesis of activation of both SPNs pathways during the learning of operant behaviors in a flexibility context for different flexibility challenges. This could set an important base on how SPNs activity changes between two different strategies, helping to further understand their role in flexibility and how is it supposed to evolve in a natural context, that is, without specific alterations in SPNs activity. In the future this might help identify the main differences between healthy and impaired flexibility specific encodings by d- and i-SPNs.