Local resources is the rocky dry land have the potential to be developed as productive land if the selection of technology and types of varieties adaptive location.  T he problem is the least information abouton the nature of agronomy and the production of  adaptive local corn varieties at rocky dryland. The study aims to : 1)  gotdata about the agronomic character of local corn varietiesat different class rocky dry land. 2) obtain data about productive corn on rocky dry land. Experiment using R andomized Block D esignwith 9 replications. Treatment is 3 varieties of maize: 1) Haikesak Putih , 2) Tastim I Putih , and 3) Bunga Merah Fatumonas. Different conditions of rocks on the surface are different locations. C riteria of a rocky dry land  class consists of Class 2: consists of  rocks that exist on the surface as much as  0.1-3.0% and the small rock is the distance is about 1 meter. Class 3 : consists of  rocks that exist on the surface as much as  3.0% -15% and the small rock is the distance is about 0,5 meter. Class 4 : consists of  rocks that exist on the surface as much as 15-25% and the small rock is the distance is about 0,3 meter. Observation variables consist of: 1)Growth: a) H eight of plant , b) Diameter of stem, c) Width of Leaf, and 2) Production: a) W eight of maize per plant ,  b) Y ield per hectare . The research results showed  that the local corn varieties of Bunga Merah Fatumonas have the best agronomic at different class of rocky dry land.  T he local corn v arieties of Bunga Merah Fatumonas provide the best response to: height of plant , weight of maize per plant , and  yield per hectare .  Key word: Rocky dry land class, best response