Hudson Volcano is one of the most active volcanoes in the southernmost Southern Andean volcanic zone, characterized by an ice-filled caldera 10 km in diameter. Tephrochronological studies indicate records of explosive activity from the late Pleistocene to historical times. In fact, the last large eruption occurred in August 1991 and is considered to be one of the largest eruptions of the twentieth century. The volcano is located in a remote and roadless region of the Patagonian Andes, which means that numerical models play an important role in assessing volcanic hazards at Hudson. In particular, these models are used to identify areas susceptible to be impacted by lahar flows and tephra fallout. In addition, a proximal-hazard zone was built using the energy cone model, which is useful when little or no prior geologic data are available. Lahar-inundation hazard zones were delineated using the LAHARZ model, based on empirical relationships. Several volumes were considered because of the range of potential lahar-initiating events, such as ice melting or mobilization of loose pyroclasts. Simulations indicate that valleys located west of the volcano are likely to be inundated by lahars, even small-volume lahars triggered by small eruptions, as have been recorded during historical episodes. In contrast, only large events would likely affect main populated settlements located farther west from the volcano. Tephra-fall deposits were simulated with an advection-diffusion model, Tephra2, employing wind data derived from atmospheric global data sets. Both spatial distribution of deposits and thickness derived from the August 1991 eruption were satisfactorily validated. Three eruptive scenarios were selected according to the geological record of the volcano. Results of simulations are outlined as probabilistic maps of mass accumulation on the surface and also as exceedance probability curves for selected localities. This analysis shows that regions east of the volcano are more vulnerable to tephra fallout throughout the year, and therefore no major interseasonal variability is recognized. However, the arrival of weather fronts, common during autumn and winter, could trigger tropospheric wind shifts, which may increase the chance of meridional (north-south) transport of pyroclasts. Finally, according to available tephrochrono-logical data, the occurrence of a large eruption was estimated, indicating 10%-20% likelihood of an eruption >= VEI 4 (where VEI is volcanic explosivity index) during the next 100 yr.