Peruvian anchovy is the largest single-species fishery of the planet, accounting for 24 to 47% of the global production of fishmeal and fish oil. Therefore, it is a crucial component of the aquaculture and farming industries that depend on these products. Peruvian anchovies live off the cast of Peru and northern Chile in the Northern Humboldt Current System (NHCS). This is the most productive eastern boundary upwelling system in the planet. In this study, we used habitat models coupled to a regional physical-biogeochemical model of the NHCS to simulate changes in the probability of occurrence of anchovies due to climate change. We used two methods to produce the habitat models: Generalised Additive Models (GAM) and XGBoost, a machine learning algorithm. The models were trained based on presence-absence observations between 1990 and 2010 using month, water depth, depth of the oxycline, sea surface temperature and distance of the coast as predictors. The XGBoost and GAM models had test balanced accuracies of 0.62 and 0.63, respectively. After training the models, we ran climate change sensitivity studies assuming a temperature increase of 2.2 °C over the whole domain. Since it is not yet clear how wind patterns will change, we ran two more experiments increasing and decreasing the wind velocity over the ocean by 10%. Our results showed a decrease in anchovies probability of occurrence of close to 60% in the XGBoost scenarios with increased temperature and decreased wind, and an increase of 40% in the scenario with higher wind speeds. The GAM model showed stronger responses than the XGBoost. Here, the probability of finding anchovies decreased by about 90% in the scenario with higher temperature and by 75% in the scenario with weaker winds while it increased by about 65% in the scenario with stronger winds. According to our study, global warming is likely to have a strong impact on the habitat of the biggest fishery of the planet.