This study evaluates the effect of an intervention whose aim is to make articulatory consciousness a tool in decoding and spelling. The sample comprises 11 students with severe dyslexia (2 "SD" below the mean pseudoword scores), and the intervention programme consists of 32 individual sessions over 8 weeks. The study applies a multiple baseline/probe design with five baseline tests that correspond to a control condition, eight tests during the intervention and five post-intervention tests. On average, the results show significant improvement in all reading and spelling outcomes. However, there were also significant effects on an irrelevant control task (the pegboard test), perhaps indicating testing effects on the dependent variables, making it difficult to draw firm conclusions from the study. Consequently, testing the intervention in randomised trials of children with severe dyslexia is recommended to draw more firm conclusions about its efficacy for this group.