Dimensional analysis was used to develop a predictive mathematical model for microwave heating of water. Water properties (density, specific heat, thermal conductivity and dielectric loss factor), power output and frequency of microwave, and geometric parameters were included in the dimensionless terms. Two household microwave cavities (donated by Tappan and Litton) were used. Time-temperature data were collected using a Luxtron fluoroptic temperature sensor (model 750) equipped with fiber optic probes. the predictive model, which was based on data taken in a Litton microwave cavity, was used to predict heating time of water in a Tappan microwave cavity. the results showed good agreement between the measured and the predicted time duration, with a high correlation coefficient (R2=. 99).