The influence of target characteristics on results of neutron-induced cross-section measurements is discussed. The basic principles of total and reaction cross-section experiments are described. The discussion shows that each application needs targets with specific requirements, which are characterized for quantities, such as the total number of nuclei per unit area, effective area and homogeneity. The result of such a characterization can have a strong impact on the total uncertainty of the quantities deduced from the measured data. Based on the measurement principles and on practical experience, recommendations for specific cross-section measurements are presented. These recommendations refer to both the target properties and to the methods used for target characterization. In addition, a characterization method based on the use of neutron resonances is presented. This method can be used to determine the presence and quantity of contaminants and impurities which have a strong impact on the results of cross-section measurements.