The quality requirements for industrial parts produced by additive manufacturing (AM), but also the safety requirements in sectors such as aeronautics or nuclear power, demand strict control across the entire value chain of the feedstock. The first stage concerns the use of a raw material that is ideally suited for the concerned technology and thus to its specific characteristics. Indeed, each machine is characterized by its specificities and its own tolerances, which means that the powders must meet criteria adapted to the targeted technology and even to the manufacturer for a given technology. The high quality of the raw material (powder) certified by the supplier is a necessary, but not sufficient, condition to guarantee the manufacture of parts without defects or with some controlled defects for reproducible properties. The powder can thus be a source of inconstancies and irregularities during successive builds and, in a deceptive manner, the resulting mechanical properties of materials can be altered and may vary over time. However, other factors such as the laser process conditions may also be responsible for poor part quality, but these are not the focus of this chapter. Thus, an extended characterization of the powders, their systematic control and their increased traceability during their lifecycle are essential steps for the qualification of AM parts. This chapter is therefore devoted to metal powders, providing a description of this raw material in relation to the specific needs of the various AM processes. The important requirements in terms of raw material quality have resulted in a standardization effort, in particular for characterization methods such as the relatively recent ASTM F3049-21 Standard Guide for Characterizing Properties of Metal Powders Used for Additive Manufacturing Processes, published jointly by the international standards organizations ASTM and ISO (ASTM 2014). The main technological challenges related to this feedstock can be summarized as follows: – a lack of fine powder characterization methods (chemistry, structure), which correspond to important input data; – a lack of clear relationships established between the intrinsic properties of the powders and the final properties of the produced materials; – a lack of understanding of the factors underlying the problems of variability from one machine to another or from one day to another for the same machine, which may be related to the reproducibility of the powder or to some machine wear (for example, the layering system); – the lack of simulation tools for powder manufacturing processes (atomization, etc.) that could be used for optimization; – the inadequacy of referring to traditional methods of qualification and certification of materials for the qualification of parts; – the still too limited qualification of metal powders, which makes it impossible, for example, to clearly identify the differences between a new powder and a recycled powder; – the still too fragmentary nature of material databases for AM with precise control and reliable traceability.