This paper deals with the extraction of the parameters of the single-diode solar cell model from experimental I-V characteristics of Si and Multi-junction solar cells. The extraction is carried out by three different optimization methods in an attempt to judge which method surpasses the others in terms of data-to-model fitting. The first and the second methods are a variation of the Newton-Raphson method and the Levenberg-Marquardt algorithm, respectively. Both methods are based on the gradient descent approach. The third method is a global-search method based on a Genetic-Algorithm. The extraction of the parameters was done in two stages. On the first stage, empirical I-V characteristics of solar cells that contained measurement errors were used, whereas on the second stage the parameters were re-extracted using a smooth synthetic I-V data. In the absence of true measured parameter values of the cells, it was left to rate the performance of the three optimization methods by the extraction error alone. Although no definitive conclusions could be drawn from the results of the noisy data, results of the smooth data are far more pronounced in terms of the extraction error, and tend to favor the Newton-Raphson method. Crown Copyright (C) 2013 Published by Elsevier B.V. All rights reserved.