In decentralized Supply Chains, each member decides based on his own interests. Conflict of interests results in suboptimal decisions and poor performance for entire supply chains, as well seriously harms credibly information sharing across them. In this thesis, coordination of decisions in supply chains in the context of Capacity Procurement problem are studied in different situations in form of three models. In first model, a dyadic supply chain with stochastic demand and exogenous price is investigated by taking various costs into account. PARD (PARtially Deductible, reservation contract) and RCRS (Reservation contract with Cost and Revenue Sharing) contracts are designed and proposed in order for coordination of decisions respectively in full and partial information updating situations. It is mathematically shown that coordination is achieved by using each contract in its corresponding situation. In second model, endogenous price is assumed. That is, demand is modelled as sum of a decreasing linear function of price and a stochastic parameter. The model is first examined in a dyadic structure, and RSRP (Revenue Sharing Reservation contract with Penalty) contract is proposed for coordinating of price, production time and production rate decisions. It is proved that coordination is achieved by RSRP contract in the dyadic structure. The application of RSRP contract is then extended to be employed in a divergent supply chain with multiple retailers, and shown that the supply chain performs considerably better than the same supply chain with a wholesale contract. In third model, a divergent supply chain comprising a supplier and multiple retailers is studied where retailers face stochastic and price-dependent demand. Since main decision makers in supply chain interactions are human, paying attention to human decision making process and their biases from theoretical predictions are important in designing coordination mechanisms. One of the non-pecuniary factors which cause deviations in human-decisions is Trust. In this model, the retailers have more accurate demand forecast information due to their proximity to market. In order to secure availability of products during the selling season, the retailers have incentives to inflate their private forecast information. A coordination mechanism is proposed, which consists of an optimization model, a scoring system and a rewarding-punishing system, in order to coordinate the supply chain. Using simulation approach, performance of the mechanism is then compared to those of two other mechanisms, namely Without Trust an Asymmetric mechanism. According to the results, employing the mechanism in situations with any demand variability is advised. More accurately, in situations with high demand variability, the mechanism achieves a proper profit improvement and moderate capability for identifying deceptive agents, while in situations with low demand variability, the mechanism shows insignificant profit improvement and considerable ability in identifying deceptive agents.