A substantial investment is made each year in research to support environmental policies. Understanding the impact of this research is important from a number of perspectives. What remains unclear is how such evaluations may be undertaken, particularly as very little current practice is captured in the literature. This paper reports on a set of 10 exploratory case studies of environmental research impact assessment in practice. Most of the impact evaluations identified have multiple objectives and used a combination of research methods. Challenges include establishing attribution, the timing of an evaluation, how to capture the duration of research impact, checking the reliability of information from key informant interviews and identifying methods for capturing as many impacts as possible. Best and Holmes' (2010) framework is used to consider the status of the case-study organisations in progressing from first generation linear models of knowledge to action to the more recently advocated systems models.