OBJECTIVE: The mouse is the most popular animal model in olfactory research. Behavior tests with odorants are essential for determining olfactory phenotype. To the best of our knowledge, the mouse olfactory behavior test has not been standardized, making the results vulnerable to inter-observer variation. We sought to develop a new mouse olfactory behavior test assessed by an automatic video tracking system with minimal inter-observer variation. METHODS: A video-tracking system was used to automatically track mouse behavior in standard breeding cages with C57BL/6N mice. We tested two odorants (peanut butter for the preference test, 2MB acid for the avoidance test) and distilled water (for a control). Mouse behavior was recorded for 3min and analyzed. For the preference test, investigation time was measured. For the avoidance test, time spent in sectors away from the odorant zone was measured. To confirm our experimental settings, we also evaluated an anosmia mouse model prepared with intranasal administration of ZnSO4. RESULTS: All strains of mice showed reproducible behavior patterns of preference or avoidance for the odorants. The anosmia mouse model, as expected, failed to show an olfactory ability for preference or avoidance, and this was well-matched by histologic changes caused by the ZnSO4 treatment. The automatic video tracking system successfully tracked and automatically calculated mouse behavior with good reproducibility. CONCLUSION: Our olfactory behavior test offers a simple and accurate method to evaluate olfactory function in mice. This test can be utilized as a possible standard method to search for features of olfactory phenotypes in mice.