Rabies is a fatal neurologic disease caused by lyssavirus infection. Bats are important natural reservoir hosts of various lyssaviruses that can be transmitted to people. The epidemiology and pathogenesis of rabies in bats are poorly understood, making it difficult to prevent zoonotic transmission. To further our understanding of lyssavirus pathogenesis in a natural bat host, an experimental model using straw-colored fruit bats (Eidolon helvum) and Lagos bat virus, an endemic lyssavirus in this species, was developed. To determine the lowest viral dose resulting in 100% productive infection, bats in five groups (four bats per group) were inoculated intramuscularly with one of five doses, ranging from 100.1 to 104.1 median tissue culture infectious dose (TCID50). More bats died due to the development of rabies after the middle dose (102.1 TCID50, 4/4 bats) than after lower (101.1, 2/4; 101.1, 2/4) or higher (103.1, 2/4; 104.1, 2/4) doses of virus. In the two highest dose groups, 4/8 bats developed rabies. Of those bats that remained healthy 3/4 bats seroconverted, suggesting that high antigen loads can trigger a strong immune response that abrogates a productive infection. In contrast, in the two lowest dose groups, 3/8 bats developed rabies, 1/8 remained healthy and seroconverted and 4/8 bats remained healthy and did not seroconvert, suggesting these doses are too low to reliably induce infection. The main lesion in all clinically affected bats was meningoencephalitis associated with lyssavirus-positive neurons. Lyssavirus antigen was detected in tongue epithelium (5/11 infected bats) rather than in salivary gland epithelium (0/11), suggesting viral excretion via the tongue. Thus, intramuscular inoculation of 102.1 TCID50 of Lagos bat virus into straw-colored fruit bats is a suitable model for lyssavirus associated bat rabies in a natural reservoir host, and can help with the investigation of lyssavirus infection dynamics in bats.