DESPITE extensive theoretical effort(1,8), the evolution of lekking as a mating system remains a controversial issue(9,10). Leks are nonresource-based matins aggregations(2), but may also be regarded as patches differing in female encounter rate(2,3,5,7). We report here a new distribution model that incorporates variation in male mating skew with lek size. The model predicts that, under specified conditions, high-ranking males have smaller optimal lek sizes than low-ranking males. All males benefit from initial clustering, but only low-ranking males gain from large aggregations. This generates progressive clustering around high-ranking males at hotspots determined by female spatial distributions. The predictions of our model were validated in two ways using empirical data on lekking ruffs, Philomachus pugnax. Our model integrates the basic elements of the previously competing hotspot(2,3) and hotshot(4) models of lek evolution by a simple mechanism, and could explain the evolution of lekking.