Knowledge of connectivity among subpopulations is fundamental in the identification of the appropriate geographical scales for stock status evaluation and management, the identification of areas with greater retention rates, and space‐based fisheries management. Here, an integration of hydrodynamic, biological, and habitat models results is used to assess connectivity and support the definition of essential fish habitats (EFH) in the Adriatic Sea, with reference to Nephrops norvegicus, an important benthic commercial resource, the recruitment of which is strongly related to larval dispersal from spawning to recruitment areas. We explored oceanographic and biological connectivity in the Adriatic Sea under a wide and representative variety of oceanographic conditions (winters 2006–2012) by tracking 3D trajectories of larvae released from different areas. We used a Lagrangian model that features a specific larval behavior module with explicit dependence on environmental parameters (i.e., temperature and sediment type) and that is driven by high‐resolution hydrodynamic and meteorological data. The results were used to partition the area in which Nephrops was observed into 20 homogenous management subareas; to assess the connection between spawning, recruitment, and harvesting grounds; and to identify potential subpopulation boundaries as well as the connectivity among the potential subpopulations. The results suggest the presence of at least three distinct subpopulations, which need to be independently managed and conserved, and confirms that the Jabuka‐Pomo pit is the most important spawning area, but alone it cannot sustain Nephrops populations throughout the Adriatic Sea. The results also show the importance to move from particle‐tracking to approaches based on integrated models.