This paper describes the context, processes and issues experienced over 5 years in which a RCT was carried out to evaluate a programme for children aged 7-8 who were struggling with their reading. Its specific aim is to illuminate questions about the design of complex teaching approaches and their evaluation using an RCT. This covers the early development by the originator and work to develop and design a RCT funded trial. The experimental, process evaluation and case studies findings are summarized. It is argued that if RCT is the only credible evaluation approach, that there is no strong evidence for IGR use. But, if RCT as the first choice evaluation approach needs to be supplemented by process evaluation, then a positive process evaluation might save IGR for further development and evaluation trials. However, it is suggested that conceptualizing IGR as a complex teaching intervention also raises questions about RCT as the method of first choice. It is argued that a Designed-Based Research approach to scaling up IGR, an example of a Design & Research approach, might have been tried. The reasons why this was not done are explored with implications for the place of RCTs in improving teaching and learning.