Nonparametric estimation methods are experiencing a growing success in their application in economics. This success comes from the fact that they do not imply any assumption about the object to estimate: probability density function, regression model… and let the data determine an appropriate model. This paper gives an introduction to nonparametric techniques used when estimating production frontier and efficiency scores from a sample of decision making units (firms, farms, …). Special attention is devoted to Data Envelopment Analysis (DEA) estimation techniques. Example are used to illustrate the implementation of these techniques.