Halahatang area is located in the northern Tarim Basin that is an important oil and gas exploration area. With the further exploration, plenty of oil and gas resources are discovered, but the complexity of distribution of oil and water, and the uncertainty of the reservoir characteristics and the main control factors seriously restricted the oil and gas exploration and development in the study area.Based on the analysis of Ha11 well block in Halahatang oilfield, we study the reservoir characteristics and main controlling factors in Halahatang area. It can be divided into two fracture-cavity belts in Ha11 well block: the Ha11 fracture-cavity belt and Ha12 fracturecavity belt. The Ha11 fracture-cavity belt is overall oil, and the drilling success rate is high with more efficient wells. But the Ha12 fracture-cavity belt is low in drilling success rate, and has a larger proportion of failure wells with abundant water. Analysis of the structure characteristics, formation characteristics, phase characteristics, fluid property variation characteristics, faults and reservoir types of the study area, we summarize the reasons accounting for the oil-water distribution differences between these 2 fracture-cavity belts in study area. The results show that the Ha11 fracture-cavity belt and the Ha12 fracture-cavity belt are similar in structure characteristic,formation characteristics and fluid property variation characteristics, and are all the same unsaturated oil reservoirs. But through the comparative analysis, the Ha11 fracture-cavity belt is mainly controlled by NW early-middle Caledonian large strike-slips and its small auxiliary faults. The Ha12 fracture-cavity belt is mainly controlled by NE early-middle Caledonian large strike-slips and its small auxiliary faults. So, that is why Ha11 well block was divided into two fracture-cavity belts, two oil-water systems. In addition, the relative level of structure and the reservoir types are also an important reason to cause these two fracture-cavity belts with different oilwater distributions.