ABSTRACT       A data matrix comprising 100 characters of morphology, anatomy and seed protein banding recorded comparatively for ten species of Cleome and Gynandropsis (Cleomaceae) was analyzed under three fundamentally different numerical methods. The PRIMER analysis used the Bray Curtis (SØrensen) distance measure together with the single linkage clustering methods. The SPSS program used Ward'sclustering as a distance measure and the average linkage clustering method distance or complete linkage distance. All three dendrograms agree in grouping the ten species into two major groups: A (including C. droserifolia and C.chrysantha) and B (including C. amblyocarpa, C. paradoxa, C. arabica, C. viscosa C. brachycarpa, C. scaposa, C. hanburyana and Gynandropsis gynandra).This indicates that Gynandropsis gynandra ought to be submerged in Cleome as Cleome gynandra