||Keywords:Market liberalisation, rural households, traders,Kenya, market integration, cointegration analysis, difference-in-difference approach, farmer response.Most countries in Sub-Saharan Africa, includingKenya, liberalised their agricultural commodity markets in the 1980s and 1990s as a strategy to increase marketing efficiency. In this thesis, we provide an account of the impacts of market liberalisation on households inNorthwestern Kenya, a maize surplus producing region. We apply several analytical frameworks including descriptive statistics, structure, conduct and performance modeling, cointegration and error correction modeling approaches, double differencing and multinomial probit regression to both primary and secondary data.A descriptive review of the market liberalisation process indicates that commodity markets are liberalised. This is particularly evidenced by the increased number of traders across the four commodity markets (fertiliser, seed, maize and milk) examined. However, the markets are only partially liberalised since there is still some government active participation in some markets like maize and milk. An analysis of the structure, conduct and performance of the four commodity markets shows that the markets are competitive. This is evidenced by low trader concentration levels and marketing margins. Furthermore, there is no evidence of collusion among traders in terms of pricing or limiting market supply despite financial and structural constraints that limit firm expansion. The ensuing increased private trader participation in commodity markets partly explains the observed marketing integration among maize wholesale markets, and a positive and strong aggregate supply response both in the short- and long-run.By comparing farming activities of maize producing households in 1992 and 2003 by using panel data, we find that on-farm diversification and commercialisation have increased. In particular, fertiliser use increased dramatically in terms of number of households using fertiliser and the rate of use. Furthermore, the number of households participating in maize marketing increased from 40 percent in 1992 to 70 percent in 2003. Area under maize, distance to the market, rate of basal fertiliser use and household size are the major factors that influence household participation/non-participation in maize markets.Overall, the study shows that market liberalisation could be beneficial to households if constraints traders face could be eliminated and that the government has lost revenue because of its reduced role in commodity marketing activities.