This thesis seeks to link the theory of assessment with the practice of dyslexia specialists in the UK. The studies outline the development of an alternative approach to the assessment of spelling in children with dyslexia that incorporates a dynamic assessment approach. The thesis presents four studies that build on each other to provide evidence for a Dynamic Assessment (DA) in the curriculum domain of spelling. The first study reports on the current assessment practices of 299 dyslexia specialists in the UK, identifying that while conventional static testing provides a measure of spelling accuracy compared to same-aged peers, further information is needed to inform intervention/support. As a foundation for an interactive (dynamic) assessment approach, Study Two builds on research by Critten et al. (2007) and Rittle-Johnson and Siegler (1999) to determine whether children with dyslexia are able to explain their spelling choices in a similar way as same-age and spelling-age matched groups (66 children aged between 6;3 and 9;9) and to investigate the spelling strategies they use. Study Three outlines the development of a Dynamic Assessment of Spelling (DASp) using the theoretical framework of Lidz (2014) for curriculum-based DA. The assessment developed for this thesis utilised a teach-train-teach format incorporating a graduated prompt procedure at post-test. Fifty children with dyslexia between the ages of 8 years and 9 years 11 months participated in this study. Information collected from the DASp procedure allowed the assessor to develop spelling profiles that revealed differences in the children's spelling abilities even when they achieved similar spelling scores on a static spelling task. Finally, Study Four reports the outcome of a pilot study with 6 qualified dyslexia specialists who used the DASp to develop an intervention programme for 12 children (mean age 8;8). All of the specialists reported improved outcomes for the children in the study and overall evaluations were positive. Implications for practice, limitations of the studies, and suggestions for future studies are discussed.