Objective. To determine any changes in tuberculosis epidemiology in the last 15 years in Benin, seasonal variations, and forecasted numbers of tuberculosis cases in the next five years. Materials and Methods. Retrospective cohort and time series study of all tuberculosis cases notified between 2000 and 2014. The "R" software version 3.2.1 (Institute for Statistics and Mathematics Vienna Austria) and the Box-Jenkins 1976 modeling approach were used for time series analysis. Results. Of 246943 presumptive cases, 54303 (22%) were diagnosed with tuberculosis. Annual notified case numbers increased, with the highest reported in 2011. New pulmonary bacteriologically confirmed tuberculosis (NPBCT) represented 78%  ± SD 2%. Retreatment cases decreased from 10% to 6% and new pulmonary clinically diagnosed cases increased from 2% to 8%. NPBCT notification rates decreased in males from 2012, in young people aged 15-34 years and in Borgou-Alibori region. There was a seasonal pattern in tuberculosis cases. Over 90% of NPBCT were HIV-tested with a stable HIV prevalence of 13%. The ARIMA best fit model predicted a decrease in tuberculosis cases finding in the next five years. Conclusion. Tuberculosis case notifications are predicted to decrease in the next five years if current passive case finding is used. Additional strategies are needed in the country.