The revised W.H.O. guidelines for malaria management in endemic countries recommend that treatment should be reserved to laboratory confirmed cases, both for adults and children. Currently the most widely used tools are rapid diagnostic tests (RDTs), that are accurate and reliable in diagnosing malaria infection. However, an infection is not necessarily a clinical malaria, and RDTs may give positive results in febrile patients who have another cause of fever. Excessive reliance on RDTs may cause overlooking potentially severe non malarial febrile illnesses (NMFI) in these cases. In countries or areas where transmission intensity remains very high, fever management in children (especially in the rainy season) should probably remain presumptive, as a test-based management may not be safe, nor cost effective. In contrast, in countries with low transmission, including those targeted for malaria elimination, RDTs are a key resource to limit unnecessary antimalarial prescription and to identify pockets of infected individuals. Research should focus on very sensitive tools for infection on one side, and on improved tools for clinical management on the other, including biomarkers of clinical malaria and/or of alternative causes of fever.