Kenyan urban system has experienced an early dominance of few major cities because of resource allocation, and colonization policies. The consequences of urban growth continued after independence, extended the gap between major cities and the rest of the urban system. That led to weak linkages among the entire urban system. This paper seeks to suggest efficient hierarchical urban system for governmental service providing with minimum cost. This is accomplished by analysing the Kenyan urban system based on the 2009 census of population and housing results. The cities potential in the urban system is evaluated, besides defining the dominating linkages among cities. This enables the illustration of urban system centralism problem in Kenya. Then a P-median clustering approach is used to propose an optimum hierarchy among cities in the system. Paper results demonstrate the flexibility of mathematical models, where the results could be re-checked as additional information became available. That is guaranteed by the use of composite weighting function. Where the use of the composite weight factor gave results closer to actual precipitation. That showed a great congruence with the results of the logical analysis of contract selection, as depicted in the selective proposed nodal hierarchy. Thus, it is suggested that policy planners need to consider as much weighting factors to avoid shortcomings in the existing urban governance policy.