In order to assist enterprises with efficient production, this paper designed and developed a production decision support system (PDSS) based on cloud computing. Cloud computing platform was the guarantee of large scale application and big data processing. The PDSS was a four-layer architecture based on MVC (Model View Controller) pattern. The model base contained various production decision models, which offered a wide range support for the realization of system decision-making function. The data layer, as a data storage medium, provided big data support for PDSS. The system design fully considered the human-computer interaction, the running was stable, and the sub-modules were well connected with each other. Therefore, the PDSS can effectively provide scientific production decision support for enterprises.