Climate indices have a long history of use to combine multi-faceted climate information for tourism resource evaluation. Traditionally, indices have been used to assess tourists’ sensitivity to destination climatic pull factors, not tourists’ sensitivity to source market climate as a push factor for seasonality-driven markets. This study addresses this gap by using tourism climate indices to assess the influence of climatic push and pull factors for seasonal fluctuations in arrivals to Antigua and Barbuda, Barbados, and Saint Lucia, from the province of Ontario, Canada (from January 2008 to December 2017). Building on the conceptual foundation of the Holiday Climate Index:Beach (HCI:Beach), this study uses an optimization algorithm to develop two indices: (1) an optimized in-situ index that estimates the climatic pull-factor of the destination, and (2) an optimized ex-situ index that estimates the climatic push-factor from the source market. Findings reveal the optimized ex-situ (push) index explains 83% (R2 = 0.830) of the variability in total monthly arrivals from Ontario and has greater predictive accuracy than the in-situ (pull) index. The research advances understanding of climatic influences on Caribbean tourism arrivals and provides the foundation for new seasonal forecast-based Climate Services (CS) for destination managers and marketers.