Hydrographic data collected during 1959 and 1960 in the Gulf of Thailand, combined with supporting climatological data, have been used to investigate the salt and heat budgets of the Gulf. A simple model including mixing, net precipitation, local river run-off and Ekman flux predicts the flushing time of the Gulf, associated with processes other than the Ekman flux, to be 12.0 months, but predicts too small a magnitude and the wrong phase for the observed seasonal salinity cycle. Adding the discharge from the Mekong River reduces the flushing time to 5.4 months and improves the fit to the magnitude of the observed salinity cycle. The phase matches if a 2-month delay in the time for the Mekong discharge to reach the Gulf is allowed. The Mekong River appears to be a vital source of fresh water to the Gulf. Using the flushing time required to balance the salt budget and a model for temperature which incorporates the climatological net heat flux values, the Gulf is predicted to be warmer than observed. The annual cycle of net heat flux required to match the observed temperature cycle has a mean and an amplitude both of which are smaller than the published climatological values, but only by about 10 W/m2 (annual net heat flux). Our simple model, using half the Mekong discharge, suggests that a uniformly distributed tracer would be flushed (90% removal) from the Gulf in 7 to 9 months, with the Ekman and non-Ekman exchange processes of approximately equal importance.