Over the last decade, scenario analysis increasingly has been used to explore the main drivers affecting global food supply and demand in the future. The aim of this study is to summarise, compare and evaluate global scenarios with a focus on global food security. We find that food security outcomes differ substantially and may be attributed to three main factors: (1) differences in scenario storylines and assumptions on drivers, (2) differences in the models that are employed to process the drivers and generate scenario outcomes, and (3) differences in the way and the extent to which results are reported. We also identify several ways how studies and models may be improved. This paper provides important and helpful suggestions for practitioners, but is also useful for policy-makers, who are usually the ultimate users of these scenario studies.