Abstract: There are more than 550 producing well in Zanarol Oilfield, of which nearly 60 percent of acid fractured wells have lost effectiveness, and need repeated stimulation. Candidate well selection is one of the key links which affects the re-fracturing effect. Based on the theory research of well selection methods, according to reservoir characteristics of Zanarol oilfield and stimulated well performance, production statistics, virtual intelligence and production prediction technique have been used to build repeated acid fracturing well selection model, a new way to select carbonate reservoirs and wells. Oilfield application shows this comprehensive method increased the effective ratio of re-stimulation, productivity and producing degree of single well have been increased by using different re-fracturing technique, playing an important role in keeping the production stable in Zanarol oilfield.