A field experiment was conducted at Makariki (Central Maluku) Experimental Farm of Maluku Research Institute for Agricultural Technology during the rainy season (July to November 2005) and the dry season (November 2005 until March 2006) that aimed to study the soil management in soybean based farming that is beneficial to be recommended on dryland. The experiment used Strip Splil Plot Design with 24 treatment combinations in three replications. Each treatment has a plot size of 6 m x 4 m. The main plots were preceding crop before soybean, which consist of: dryland rice (Cl) and corn (C2), sub plot were soil tillage systems, which consist of: no tillage (TO),minimum tillage (T2), and conventional/intensive tillage (T3), and sub of sub plot were alternative fertilizers, which consist of: no fertilizer (FO), rhizoplus microbe fertilizer (Fl), cow manure (F2), and NPK inorganic fertilizer (F3). Data were analysed through qualitative and quantitative with econometric approach. The results showed that rice-soybeau rotation was more efficient in production cust compared to com-soybean rotation. The use of inorganic fertilizer (NPK) on soybean after rice on dryland is not recommended, because the marginal economic benefit was dominated by that without fertilizer as well as that with alternative fertilizer (rhizoplus). Returning rice straw residues that was followed by intensive soil tillage and rhizoplus (200 g 11a-1 )could be recommended as sultable soil management system for soybean based farming on dryland. (by :   Alfons.J.B; Hedayana. R.  Dalam jurnal Budidaya Pertanian  Vol 6(1)  hal. 30-38)