Given the worldwide increase in energy demand and its high biomass potential, sugarcane is more and more valued as an energy crop. To cope with this demand, sugarcane needs adapted cropping systems, including cultivars, and requires decision tools such as crop growth models fitted onto energy outputs (aerial dry biomass) from appropriate cultivars. To investigate biomass-energy potentialities and adapt the cane crop growth model Mosicas, experiments comparing the same energy canes cultivars from WICSCBS (Barbados) against local check cultivars were conducted on plant crops in two experiments in two contrasting sites in Réunion (Indian Ocean) and in three experiments at one site in Guadeloupe (Caribbean). Dry biomass and water content were monitored on subplots to determine biomass potentialities and select the best cultivars. Together, the plant-crop results of the two sites in Réunion were used to calibrate two variables in Mosicas: aerial dry biomass and water content. Plant-crop results from Guadeloupe were used to evaluate the model on those two variables. At the three sites, the different temperatures, radiation and rainfall patterns lead to different dynamics of cane water content and dry biomasses and production of 75 t cane/ ha. At the two sites in Réunion, using four and three parameters, respectively, for dry biomass and water content calculations, Mosicas predicts those two variables fairly well with respective root mean square errors (rmse) of 2 to 3.6 t/ha and l.1 to 1.9% according to cultivars. In Guadeloupe, the model fitted in Réunion predicted reasonably well dry biomass and partially the water content. When only the parameter values are fitted and empirical water content process are added, the modified Mosicas model predicts very different biomass dynamics and cultivar ranks. Given this, Mosicas could be used as decision-making tool to optimize energy production strategies using long-term climate data or future weather scenarios.