There is now a strong evidence base from theory and research providing a "template" to inform practice at Wave 2, guiding the design and implementation of time-limited effective early intervention programmes for pupils identified as "at risk" of reading difficulties following initial literacy instruction (Rose, 2009). In contrast, there is currently an absence of evidence to guide the more fine-grained selection and design of specific interventions (Wave 3) for those nonresponders to otherwise effective reading intervention. In this paper, we first outline our current understanding of the characteristics of effective early intervention programmes, and of the pupils who do not respond. Three theoretically motivated single-case studies from developmental theory and processing models of skilled performance are then considered as a source of evidence providing useful insights into the type of assessment needed to inform the planning of highly individualised intervention programmes for pupils with severe and persisting literacy difficulties. (Contains 2 figures.)