Land cultivation for agricultural purposes in semiarid area is usually carried out only once a year specifically during the rainy season. The condition is even worse since it is not without the risk of failure because of dry-spell or water-logging. To cope with this situation, the researchers developed a model of Rainwater Management for Agriculture Decision Supporting System (RMA-DSS). The objective of this RMA-DSS is to facilitate the decision making to build water infrastructure. Using this program it is hoped that sufficient water supply for specific crops with correct planting time can be guaranteed, which in turn will optimize harvest. The model consists of three parts, namely, rainfall-runoff-infiltration model, crop water requirement-irrigation-drainage model and rainwater management for agriculture model. The Models are designed using Microsoft Excel?s Macro Visual Basic and finalized with Visual Basic language program for operating spatial database of map object and non spatial database.