Protein peak spectrum assignment is a prerequisite of the nuclear magnetic resonance study of a molecule. We present here a computer tool which proposes the determination of the amino acid type from the values of the chemical shifts. This tool is based on two consensus algorithms based on several published typing algorithms and was trained and extensively tested against the Biological Magnetic Resonance Bank chemical shift data bank. The first one accomplishes the analysis with support vector machine technology, grouping related amino acids together, and presents a mean rate of success above 90% on the test set. The second one uses a classical consensus algorithm of vote. Furthermore, secondary structural prediction is available. This tool can be used for assisting manual assignment of peptides and proteins and can also be used as a step in an automated approach to assignment. This program has been called CRAACK and is publicly available at the following URL: http://abcis.cbs.cnrs.fr/craack.