This paper presents mathematical models of thin layer forced convection solar drying of Cuminum cyminum using two drying methods (mixed and indirect) at different operating conditions. The average initial moisture content of the seeds for all tests was about 43% d.b. and the drying was performed continuously, in each test, for a uniform period of 90 minutes drying time in a solar cabinet dryer to obtain an average final moisture content of 8% d.b. Three airflow rates (0.084, 0.127 and 0.155 m3 s-1) were adopted and the experiments were run each sunny day from 11:30 to 13:00 with an average solar intensity of 750 W m-2 (±50 W m-2), ambient air temperature of 27°C (±1°C) and relative humidity of 30% (±1%). In order to find the most suitable form of thin layer solar drying model, eleven different mathematical models were selected using the experimental data to determine the pertinent coefficients for each model by applying the non-linear regression analysis technique. The goodness of fit was evaluated by calculating and comparing the statistical values of the coefficient of determination (R2), reduced chi-square (c2) and root of mean square error (RMSE) for any model and for the two drying methods. The best results were found for the approximation of diffusion model with R2= 0.995, c2=0.0023 and RSME= 0.0199 in mixed mode type, and the Midilli model with R2= 0.994, c2=0.0045 and RSME= 0.0225 in indirect mode type thin layer solar drying.