Examples from both marine and terrestrial systems have supported the hypothesis that predation is higher in tropical than in temperate habitats and that, as a consequence, tropical species have evolved more effective defenses to deter predators. Although this hypothesis was first proposed for marine sponges over 25 years ago, our study provides the first experimental test of latitudinal differences in the effectiveness of sponge chemical defenses. We collected 20 common sponge species belonging to 14 genera from tropical Guam and temperate Northeast Spanish coasts (Indo-Pacific and Mediterranean biogeographic areas) and conducted field-based feeding experiments with large and small fish predators in both geographic areas. We use the term global deterrence to describe the deterrent activity of a sponge extract against all of the predators used in our experiments and to test the hypothesis that sponges from Guam are chemically better defended than their Mediterranean counterparts. Sympatric and allopatric deterrence refer to the average deterrent activity of a sponge against sympatric or allopatric predators. All of the sponges investigated in this study showed deterrent properties against some predators. However, 35% of the sponge species were deterrent in at least one but not in all the experiments, supporting the idea that predators can respond to chemical defenses in a species-specific manner. Tropical and temperate sponges have comparable global, sympatric, and allopatric deterrence, suggesting not only that chemical defenses from tropical and temperate sponges are equally strong but also that they are equally effective against sympatric and allopatric predators. Rather than supporting geographic trends in the production of chemical defenses, our data suggest a recurrent selection for chemical defenses in sponges as a general life-history strategy.