This study was initiated in western Shewa and eastern Wellega zones of Oromia Regional State to describe the seed system and assess the effectiveness of the seed testing and release mechanism; identify how farmers acquire and transfer maize seed; explore problems related to farmers’ seed acquisition and transfer mechanisms; and document the use of released maize varieties and hybrids. The major maize growing areas of western Oromia were purposively classified into four strata: Bako-Tibe and Chaliya Woredas in western Shewa, and Bila-Sayo and Sibu-Sire Woredas in eastern Wellega. Five peasant associations (PAs) were selected from each stratum. From each PA, 11 farmers were randomly selected and interviewed using a structured questionnaire. The total sample size was 220 farmers. Descriptive statistics and tobit analysis were used to assess farmers’ adoption of improved maize seed and their seed management practices and strategies. Total area under improved maize increased in the study area from 1992 to 1996 and total area under local varieties decreased. Since 1995, total production of improved maize has surpassed the total production of local varieties. During the 1996 cropping season, most farmers planted seed that had been saved from their previous harvest. The Ministry of Agriculture was the next most common seed source. In all woredas, BH-660 was the preferred maize cultivar for its high yield, seed size, and early maturity. The main problems constraining the use of improved maize seed were its high price, unavailability, and sometimes the distance traveled to acquire improved seed. All farmers in Bako-Tibe and Sibu-Sire and about 83% of farmers in Bako-Tibe and Bila-Sayo replaced their hybrids every year. The average time since farmers had purchased hybrid seed was 1.4 years in Chaliya, 2.3 in Bako-Tibe, 2.1 in Bila-Sayo, and 2 in Sibu-Sire. Most farmers selected seed during harvesting using good looking grain/cobs as the most important selection criterion. Seed quality was good, according to most farmers, and was mostly judged on the basis of grain filling and germination rate. Farmer education, extension services, off-farm income, and livestock units all significantly and positively influenced the area of land allocated to improved maize. A farmer’s proximity to a formal seed source had a positive impact on the allocation of land to improved maize. Next to the extension system, farmer-to-farmer seed exchange remained the most important mechanism for disseminating improved maize technology. The government can strengthen this informal system through the provision of low-interest finance, technical assistance, and publicly bred parent material. The extension system should advise farmers on the characteristics and correct adaptation zones of newly released maize cultivars. Maize breeding programs need to give proportionate weights to yield and non-yield characteristics when selecting the best varieties.