AIM: To simulate and compare accommodation in accommodative and non-accommodative human eye models. METHODS: Ray tracing and optical design program was used. Three eye models were designed and studied: the Navarro, the Arizona and the Liou-Brennan. In order to make the Navarro and Liou-Brennan models to accommodate, specific geometric parameters of the models were altered with values that were chosen from the literature. For the Arizona model, its’ mathematical functions for accommodation were used for the same accommodative demands. The simulation included four distances of accommodation for each model: at infinity, 3, 1 and 0.5 m.The results were diffraction images of a “letter F” for graphical comparison, spot diagrams on the retinal field and Modulation Transfer Function (MTF) graphs. RESULTS:Zernike coefficients for the aberrations, Airy disk diameter, root mean square (RMS) error diameter and total axial length of the model were provided from the program. These were compared between them in all distances. The Navarro model had the smallest axial length change as a simple model. The Arizona did not change its axial length because it is designed to be accommodative. The Liou-Brennan model had different results concerning the aberrations because of the decentration of the pupil. The MTF graphs showed small differences between the models because of the differences in their designs. CONCLUSION: All the three models are able to simulate accommodation with the expected results. There is no model that can be assumed as the best choice. Accommodation can be simulated in non-accommodativemodels and in customized ones.