Collective drip irrigation (CDI) technology is becoming increasingly important for saving water and fertilizer and for improving land productivity. However, many farmers stop using this technology after investing in it and using it for several years in China. Sizable transaction costs occurring in CDI management may play a major role in CDI usage, but little research has been conducted to date. This paper measured transaction costs in traditional and new methods. Empirical estimates of the effects were obtained by applying Heckprobit and probit models to data collected in Awat County, Xinjiang, China, in 2017. We find that (1) new transaction cost proxies related to employing/demising irrigators and checking outcomes of irrigation and fertigation have significantly positive effects on the use decision of CDI and that (2) the traditional transaction cost proxy related to group size has a U-shaped relationship with the probability of using CDI, while the proxy related to social heterogeneity has significantly negative effects on this probability. These findings suggest that transaction costs originating from uncertainty are an important factor in limiting the use of this technology. Additionally, this paper contributes to the literature by examining the relationship between transaction costs and collective action.