This paper proposes production system, which is one of the most typical structures in artificial intelligence, to solve the puzzle in drilling fluid design. It is the first time that a kind of rule storage technique is put forward, which is easy in implementation and can settle the composition issue between the antecedent and consequent of rules in productions. The core of production system—the representation of knowledge, rule-based reasoning mechanism and conflict resolution in reasoning processes—is presented. According to analyzing such specific data as the geological significance and the geometry of well in a certain stage of oil-well, the rational type of drilling fluid is devised to compensate for the shortcoming of manual design and to optimize the drilling, meanwhile, scientific performance parameters of drilling fluid are furnished.