The study compares presence-absence and presence-only methods for predicting bird habitat suitability. Using breeding bird atlas data from Catalonia, the authors evaluated the performance of Ecological Niche Factor Analysis (ENFA) with presence data only and Generalized Linear Models (GLM) with presence/absence data. Results showed that GLM predictions were more accurate than ENFA, especially when absence data was reliable. Species with more restricted ecological niches were modelled less accurately, and models for wide-ranging species were more sensitive to absence data. The study emphasizes that species ecological characteristics significantly influence model accuracy, and presence/absence methods are particularly important for predicting distributions of generalist species. It concludes that presence/absence methods should be preferred when absence data is available, as they provide more accurate predictions. The study also highlights the importance of considering species' ecological niche and prevalence in model accuracy, and that habitat suitability models should be tailored to specific applications. The results suggest that presence/absence methods are generally more accurate than presence-only methods, especially for species with more restricted niches. The study underscores the need for careful consideration of data quality and species characteristics when selecting habitat suitability modelling methods.The study compares presence-absence and presence-only methods for predicting bird habitat suitability. Using breeding bird atlas data from Catalonia, the authors evaluated the performance of Ecological Niche Factor Analysis (ENFA) with presence data only and Generalized Linear Models (GLM) with presence/absence data. Results showed that GLM predictions were more accurate than ENFA, especially when absence data was reliable. Species with more restricted ecological niches were modelled less accurately, and models for wide-ranging species were more sensitive to absence data. The study emphasizes that species ecological characteristics significantly influence model accuracy, and presence/absence methods are particularly important for predicting distributions of generalist species. It concludes that presence/absence methods should be preferred when absence data is available, as they provide more accurate predictions. The study also highlights the importance of considering species' ecological niche and prevalence in model accuracy, and that habitat suitability models should be tailored to specific applications. The results suggest that presence/absence methods are generally more accurate than presence-only methods, especially for species with more restricted niches. The study underscores the need for careful consideration of data quality and species characteristics when selecting habitat suitability modelling methods.