In previous studies, we have demonstrated that a Sniff Olfactometer can be used to determine a human subject’s recognition threshold of an odorant using six different concentrations. These concentrations were carefully selected to have a noticeable difference between them, and they were arranged in two triads with alternating orders. We conducted experiments where we increased the time gap between the presentation of the two triads from one minute to five days, and we found that this time gap did not disrupt the threshold function. This suggested that when people smell an odorant, their brain does not process the stimuli individually but rather compares them. To further investigate the influence of prior stimulation on odor recognition of odorants near the threshold, we initiated another project based on the Bayesian model. In Bayesian analysis, prior beliefs or knowledge about the parameters of a statistical model are combined with new data to form a posterior distribution, which is then used to make predictions or decisions. In our study, we designed a sequence of low, medium, and high concentrations to establish known relationships between the concentration of the prior stimulus and the posterior stimulus. Through multiple tests with different human subjects, we observed that the probability of perceiving the posterior stimulus changed relative to the concentration of the prior stimulus. Specifically, a higher concentration prior was associated with a lower probability of perceiving the posterior stimulus, and vice versa. By demonstrating the Bayesian rule in odor perception, our findings have potential real-life applications. For instance, in the hospitality industry, hotels and restaurants could create specific odorant environments to enhance customers’ experiences with food or living. The prior stimuli, such as pleasant or appetizing scents, could be strategically introduced to influence the perception of subsequent odors and create a more enjoyable atmosphere.