Abstract Flamanville marine protected area (MPA) located in Normandy, France, was created in 2000 to sustain the local crustacean fishery. In 1985, an annual survey targeting crustacean populations (e.g. European lobster Homarus gammarus, edible crab Cancer pagurus, and European spider crab Maja brachydactyla) that include the future MPA started. The MPA implementation effects were investigated in this study. The sampling design allowed the integration of spatial and temporal covariances to estimate the MPA effects. With respect to the initial objective, the MPA significantly improves the most economically valuable species, H. gammarus, abundance. Lobster catch per unit effort increased by 597% in the MPA, whereas outside it only increased by 156%. The MPA creation also led to an increase in lobster size inside the protected area. Furthermore, a few years after the MPA implementation, lobster catches showed a significant non-linear decline with distance from the centre of the reserve, suggesting a spillover effect. However, the edible crab catches were halved following the MPA implementation suggesting an opposite effect. Spider crab abundance seemed unaffected. Finally, the MPA implementation had no effect on edible crab and spider crab sizes. These differential responses appear to result from these species’ variable movement and competitive abilities.