The ability of black and brown lemurs (Eulemur macaco and Eulemur fulvus) to make inferences about hidden food was tested using the same paradigm as in Call's (J Comp Psycol 118:232-241, 2004) cup task experiment. When provided with either visual or auditory information about the content of two boxes (one empty, one baited), lemurs performed better in the auditory condition than in the visual one. When provided with visual or auditory information only about the empty box, one subject out of four was above chance in the auditory condition, implying inferential reasoning. No subject was successful in the visual condition. This study reveals that (1) lemurs are capable of inferential reasoning by exclusion and (2) lemurs make better use of auditory than visual information. The results are compared with the performances recorded in apes and monkeys under the same paradigm.