Fine flavor cocoa quality is significantly impacted by its postharvest drying process which can be enhanced through mathematical modeling. The objective of this study was to carry out the mathematical modeling of cocoa bean drying kinetics for a greenhouse dryer and considering intermittent drying. In this study, a natural convection solar greenhouse dryer was used to dry cocoa beans with negligible losses to an acceptable moisture content (6.6%) and cotyledon pH (5.25). Models based on Fick’s second law of diffusion, Newton’s model, and Hii, Law and Cloke’s model were developed considering (i) the entire drying period, and (ii) intermittent drying for Fick’s and Newton’s. Effective diffusivities and drying constants increased up to 1.01 ×\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ imes$$\\end{document} 10–11 m2/s and 0.0165 1/h respectively during drying. Newton’s model slightly outperformed Fick’s model, and the models considering intermittency showed slight improvement over the models for the entire drying period. Considering intermittency when modeling cocoa drying may be useful since a better fit to experimental data resulted, and variation of effective diffusivity and the drying constant was observed over the drying period.