Grain & Graze was a mixed-farming systems program conducted across nine regions of Australia over 2003–08. It had a goal of ensuring adoption of recommended practices on 6800 farms within the 5-year life of the program. This extension-based success target was further reflected in adoption targets set in contracts for each of the nine regions, and embedded into the program’s extension initiative, the Change-on-farm strategy. By 2008, the program had achieved adoption on 3200 farms. While less than half the target, this was considered by many a remarkable achievement, raising questions about the efficacy of adoption targets as a measurement of success. In a program based on devolution, regional delivery and local empowerment, the targets were contested between participants on other grounds. This paper explores how the targets were set, what Change-on-farm supported, what it achieved and how its success related to adoption targets. Using the Grain & Graze program as a case study, the paper concludes that the notion of targets as a motivator of success rather than as a measure of success is pertinent in complex systems-based research and development. The authors do not advocate avoiding targets, but suggest that both targets and the evaluation process by which success is measured be mutually negotiated in the true spirit of participatory process.