McGill University Library has received thousands of daily visitors prior to COVID-19, making it difficult sometimes for students to find seats. A limited number of seats has been made available during the pandemic to follow public health safety guidelines, with seat reservations required beforehand. While the daily number of reservations and visitors is being used to calculate attendance and space occupancy rates, this informs decision-making about service hours and expansion of seats based on descriptive statistics of past usage data that does not consider future demand. The Poisson distribution was investigated to determine whether future library space usage could be accurately predicted based on prior usage and, if true, for how long the prediction would last. This poster will present findings and lessons learned from applying the Poisson distribution to predict library space usage as well as offer a comparison of results from working with other discrete probability distributions.