Considering a twoechelon supply chain with one manufacturer and one retailer ，this paper analyzes the joint optimization of price, order lot size and production batch facing price-elastic demand D(p)=αp-k(k>1). Normally the decision models of supply chain are highly nonlinear, and it is hard to use conventional nonlinear programming method to solve them, so we propose geometric programming. Finally, a numerical example is illustrated to indicate the correctness of the solutions, which also demonstrates that joint decisions could not only improve holistic benefit of supply chain, but also improve social utility.