In many cities of developing countries, such as Addis Ababa, water supply service is poor. At the moment, Addis Ababa city which accounts for over a quarter of the nations urban population is facing unreliable and inadequate supply of water and sanitation services. The paper raises crucial issues regarding the necessity to better understand the demand for improved water supply service in the absence of market and tries to link in relation to customers’ participation in the provision of improved water supply service. In an attempt to examine the determinants of the value of improved water service and to establish how much consumers are willing to pay, we use a contingent valuation method (CVM). The tobit model shows socio-economic and demographic variables such as Income of the household, sex of the respondent dummy (female = 1), Family size, education (both secondary and tertiary level), households' year of stay, households not using tank as a storage, wealth of a household, employment status of the respondent dummy (employed = 1) , households satisfaction with the existing service dummy (unsatisfied=1), and location of the study site (Addis Ketema) are significant variables that explain the willingness to pay. The mean willingness to pay (WTP) values are 15.34 and 20.20 cents per bucket (a 20 liter container) above the exiting average tariff of 7.14 cents per bucket as calculated from the tobit and probit model, respectively. The empirical evidence is an important input for strategic planning to ensure that improvements proposed potentially improve cost recovery initiative and increase the level of customer satisfaction.