This work evaluated a collection of hundred groundnut (Arachis hypogaea L.) varieties from different origin using twenty four (24) agro-morphological traits that can help to enhance selection efficiency in crop improvement. The experiment was carried out at the experimental station of INRAN-Tarna, in the region of Maradi (Niger) during the rainy season of 2010. Analysis of variance showed a large variability among varieties for the agro-morphological traits. Principal Component Analysis (PCA), Agglomerative Hierarchical Clustering (AHC) and Fisher Discriminant Analysis (FDA) revealed that this variability is structured into four distinct groups. Groups I and II consisted of early varieties that have a high emergence rate and high pods and seed weight. These groups included mainly local varieties and those introduced in Niger through seed dissemination. Groups III and IV are composed of late varieties with large pods while group III had mostly varieties with long leaflets. Understanding the genetic control of the most discriminating among the studied traits would bring significant contribution to the genetic improvement of this important crop.