Mangosteen (Garcinia mangostana L.) is one of the economically important fruits of Thailand. Recent studies show that the climatic variability affects the flowering period of tropical fruit trees. The objectives of this study are: 1) to investigate the correlation between climatic factors (in particular with, rainfall, maximum temperature, minimum temperature, relative humidity, and dry period before flowering) and mangosteen flowering date and 2) to develop the poisson regression model to predict the flowering date of mangosteen in Nakhon Si Thammarat Province, Thailand. This model is useful in guiding farmers to manage mangosteen orchards. The results showed that the rainfall, the maximum and minimum temperatures, the relative humidity, and the dry period before flowering dates affect mangosteen flowering dates. The model which will be used as a guideline for mangosteen flowering date prediction was log(μ)=10.85+0.0001x1-0.0564x2-0.0634x3-0.0232x4+0.0003x5 where μ where is the mean of mangosteen flowering date, x1 is rainfall, x2 is maximum temperature, x3 is minimum temperature, x4 is relative humidity, and x5 is dry period before flowering date.