Most agricultural products are harvested at high moisture content that is not proper to storage. Therefore, to prevent deterioration, drying operation is used. Heat pump dryers have been widely considered for drying fruits, vegetables, and biological materials due to the low power consumption, high efficiency, high quality of final product, good consistency with environmental laws, ability to control the drying process, as well as being independent of the environmental conditions. The objectives of this study were investigation the drying kinetics and qualitative parameters of tomato slices during heat pump drying operation, empirical and analytical modeling of the process, as well as assessing the heat pump system performance. Fresh tomato samples (Shanoon variety) were purchased from a local greenhouse in Isfahan (central Iran). Heat pump drying and hot air drying (conventional method) were conducted at three temperatures 40, 50, and 60 °C and two air flow velocities 1 and 1.5 m s -1 with three replications. The heat pump system consisted of two condensers, a compressor, an expansion valve, and two evaporators, which were connected to each other using copper tubes. The refrigerant A404 was chosen for working fluid. The color indices (L*, a* and b*) were measured by a digital camera and Photoshop software. Volume change caused by shrinkage was determined using the displacement of toluene with density of 0.87 g cm -3 . To modeling mass transfer of tomato slices, the equation of Fick’s second law for infinite plane sheet was explicitly solved. Empirical modeling of drying data was also done by fitting 12 commonly used empirical and semi-empirical thin layer drying models. According to the drying curves, the constant rate period was not observed for all treatments and the drying kinetics of the samples occurred continuously. Drying time with heat pump was shorter than that of conventional method. The minimum and maximum values of effective moisture diffusion coefficient of dried tomato slices in the heat pump dryer were 3.04001×10 -9 m 2 s -1 at the temperature of 40 °C and air flow velocityof 1 m s -1 , and 6.18135×10 -9 m 2 s -1 at the temperature of 60 °C and air flow velocity of 1.5 m s -1 , respectively. In addition, maximum and minimum values of the effective moisture diffusion coefficient of the samples dried with hot air dryer were 2.53334×10 -9 m 2 s -1 at the temperature of 40 °C and air flow velocity of 1 m s -1 , and 5.26935×10 -9 m 2 s -1 at the temperature of 60 °C and air flow velocity of 1.5 m s -1 , respectively. The prediction of models based on the explicit solutiuon of the equation of the second Fickian law showed a good correlation with the experimental data at all drying temperatures and air flow velocity for both methods. According to the maximum coefficient of determination (R 2 ) and minimum value of root mean square of error (RMSE) between the observed and predicted moisture ratios, the diffusion approach model indicated the close prediction compared to the observed values. With decreasing moisture content, the color parameters, (L*, a* and b*) decreased and the shrinkage increased for all treatments. The coefficient of performance (COP) of the heat pump system increased with increasing air temperature and air flow velocity. The maximum and minimum values of COP were 5.77 and 2.21 at the temperature of 60 °C and air flow velocity of 1.5 m s -1 , and at the temperature of 50 °C and air flow velocity of 1 m s -1 , respectively. Keywords: Coeffecint of perfomance, Energy consumption, Heat pump, Mathematical modeling, Shrinkage