This paper deals with investigations carried out since 1994 on the hydrogeology of the Gran Sasso Massif (Central Italy). The Gran Sasso aquifer is a karst-partitioned aquifer of about 700 km <sup>2</sup> with well defined hydrogeological boundaries (Celico, 1983; Boni et alii, 1986). It is one of the most representative karst aquifers of the Apennines, for the following reasons: i) it contains huge groundwater resources usable for human purposes; ii) in the Gran Sasso Massif, there is a significant interaction between groundwater and underground works (tunnels and underground lab--Monjoie, 1980); iii) the Massif hosts protected areas (natural parks and wetlands) with high environmental value. The Gran Sasso hydrogeological system is composed of Meso-Cenozoic carbonate units (aquifer) and is bounded by terrigenous units represented by Miocene flysch (regional aquiclude) along its northern side, and by Quaternary continental deposits (regional aquitard), along its southern side. The Gran Sasso karst aquifer contains a unique regional-wide groundwater table (hydraulic gradient: 5-20 per thousand). This groundwater feeds springs at its border with high discharge values (more than 1 m <sup>3</sup> s (super -1) each), located at low altitude along its southern side. The groundwater also feeds other springs (mean discharge: 0.1-0.5 m <sup>3</sup> s (super -1) each), located at high altitude along its northern and eastern sides. The total discharge of the Gran Sasso springs is about 25 m <sup>3</sup> s (super -1) , corresponding to a net infiltration of more than 700 mm/year (Massoli-Novelli & Petitta, 1998). The study reported in this paper was based on previous investigations on the Gran Sasso Massif which were conducted up to the 1980s. The study had the following aims: i) to verify the location of, and to evaluate, the total groundwater resources, and also the through water budget estimation; ii) to acquire new and detailed data about the different spring groups; ii) to check the long-term hydrological effect of the drainage caused by underground works (highway tunnels and Nuclear Physics laboratory) on the discharges of the Gran Sasso springs. iii) to improve the hydrodynamic model of the Gran Sasso aquifer, considering groundwater flowpaths and interactions between adjacent hydrogeological systems. In a karst aquifer, groundwater flowpaths depend on many factors (structural setting, karst evolution, vadose zone, local recharge effects, geological framework near the springs). The sum of these factors affects groundwater hydrodynamics, flowpaths and hydrochemistry, causing a specific distribution of water resources (Cook & Solomon, 1997; Emblanch et alii, 1998; Lastenet et alii, 1995; Mudry, 1987; Scanlon et alii, 2002). Taking into account all these factors we collected and globally analysed several hydrological, hydrogeological and hydrochemical data which validated our assumptions or enabled us to formulate new assumptions on groundwater circulation. This effort resulted in a new hydrodynamic scheme of the groundwater flow of the aquifer (fig. 1), explained by the attached hydrogeological map. Discharge measurements of springs and rivers showed ground-water decrease in comparison with the discharges before the highway tunnel drainage (tables 1 and 3, figs. 2, 3 and 6). During the past few years (1996-2000), discharges showed a moderate increase. The decrease is due not only to the tunnel drainage, but also to climate variations (fig. 18 and 19). Groundwater level monitoring at the borders of the karst aquifer helped to elucidate the relationships between the karst aquifers and the Quaternary alluvial aquifers and aquitards. In the L'Aquila Plain (fig. 1), groundwater flows from the Gran Sasso aquifer to the multilayer plain aquifer, to its main springs and to the Aterno River (figs. 4, 7 and 8). In the Tirino Valley (fig. 1), the low permeability of the Quaternary alluvial deposits does not allow the same transfer (fig. 5 and 9). Statistical analysis of chemical-physical spring parameters showed the difference between springs fed by local and perched aquifers and springs fed by the regional aquifer, confirming our hypothesis on the main groundwater flowpaths (table 2, figs. 10 and 11). The differences between the six spring groups recognized in the Gran Sasso Massif from the hydrochemical study (table 4, figs. 12, 14, 15 and 16) support the assumption that the groundwater flows from the core to the boundaries of the massif, depending on the structural setting, the karst development and evolution, the local recharge effects and on the geological framework near the spring. Our hypotheses on the recharge area of the main springs were confirmed or refuted by isotope analyses (figs. 13 and 17), which were only used to integrate the previous hydrogeological model, based on field surveys. In the hydrogeological model proposed in this paper, the groundwater runs from its recharge areas (concentrated at the core of the aquifer) to the springs, following the sequence from the centre (S45-S46-S47, table 2) to the boundaries (S21-S22-S10-S11), and down to the remote springs (from S35 to S44). Along the way, salinity and temperature increase, complicated by the structural and geological setting. Many faults and thrusts (fig. 1) cause local groundwater divides, although the groundwater frequently flows between adjacent sectors. Groundwater seepages are observed both to and from the boundary aquifers. The findings from this study will be used to develop a better management strategy of the groundwater resources, taking into account the planned new underground works and the need for environmental protection in a National Park area.