The objective of this study was to investigate the effect on Mullus barbatus (red mullet) codend selectivity by increasing the codend mesh size or reducing the number of meshes in the codend circumference. This was performed through experimental sea trials and computer simulations. The sea trials were carried out in the Adriatic Sea using a polyamide codend in order to assess its selectivity, to record morphological (height, width and length) and population size structure parameters of red mullet (Mullus barbatus). The selectivity process in the codend was also simulated with an individual-based model (PRESEMO) and a finite element model (FEMNET) to calculate the codend shapes. By adjusting the behavioural description of red mullet in the simulation, the model results reached an agreement with the experimental results. Next these, behavioural parameters were used in the simulation model to predict the effect of changing mesh size and number of meshes around in new codend designs. The predicted effect and the benefit of enforcing these designs on catch efficiency below and above minimum landing size (MLS) were investigated. The results confirm that the reduction of number of meshes in the codend circumference will be more beneficial than increasing the mesh size.