Abstract: In recent decades, as crop production has increased in many areas where irrigation projects have been implemented, the global agricultural development community has promoted irrigation investments. However, due to the disappointing performance of irrigation farming in developing countries, irrigation intervention in Africa South of the Sahara including Ethiopia is an issue of debate. Moreover, several gaps exist in the Ethiopian irrigation farming literature. For instance, evidence about the direct and indirect effects of irrigation water on agriculture is not well documented. The irrigation farming literature has not disentangled the indirect effects of having access to irrigation water from the direct effect and the indirect effects have been underrepresented. Furthermore, most previous studies have applied either a quantitative or qualitative approach and have relied only on revealed data as main type of methodology, making studies that combine qualitative and quantitative research and that use both stated and revealed data underrepresented. In this study, different approaches have been applied to investigate how irrigation water impacts Ethiopia agriculture with special attention being given to disentangling the direct and indirect effects of irrigation water on Ethiopian agriculture. Using a structural equation model, a stochastic production frontier approach, and a discrete choice experiment, I drew evidence regarding the direct and indirect effects of irrigation water on crop revenue of smallholder farmers, the technical efficiency of irrigation user farmers, and the farmers’ willingness to pay to improve poor irrigation schemes from field observations, semi-structured interviews and focus group discussions with farmers, and key informant interviews with the local agricultural agents from the Koga and Fogera Districts of Amhara Region Ethiopia. The results indicate that irrigation water in general has both direct and indirect positive effects on agriculture, and the indirect effect is mediated by both improved farm inputs and the type of crops produced. The results also show that – due to poor extension services and backward agronomic practices, the mean technical efficiency of farmers in Ethiopia is very low, and that large-scale irrigation users are less technically efficient than small-scale irrigation users. Moreover, the results show that improving irrigation schemes shifts the frontier up, and smallholder farmers are strongly willing to contribute financially to the maintenance costs of irrigation schemes. The results offer relevant lessons for policymakers that providing irrigation water supply must be embedded in a comprehensive support package including access to extension services, improved input supply, and access to stable markets.