This thesis evaluated the first three years of the implementation of Reading Recovery (RR) in a Local Education Authority in the North West. This intervention aims to bring the lowest achieving NC Year 1 and NC Year 2 children to within the average attainment band for their class group. In the current education climate targeting early reading and the initiatives of the National Literacy and the National Year of Reading, this study targets early reading and builds on substantial existing research. It aims to contribute to the depiction of an effective early intervention model for reading difficulties. With this aim, the study looked at the effects of the programme on (a) the pupils' outcomes of achievement, and (b) the perceptions of the children, headteachers and RR teachers regarding the programme. In terms of the outcomes of achievement, the first cohort of the children, drawn from nine schools, who participated in the intervention was screened and formed the Reading Recovery Group (RG, N=35). They were matched with a Comparison Group (CG, N=32) from the same class, for age and gender. The groups were assessed on four occasions (T1- T4) using Clay's Observation Survey (OS), the Carver Word Recognition Test and the Salford Reading Test. A group with average attainment was also included in the comparison (being one standard deviation above or below the mean attainment of the class) and was assessed at T1 and T4. Group and Time of testing, gender, free school meals and pre-school experience, were used as the independent variables to test specified hypotheses related to scores on the dependent variables of the above assessments, and the duration (number of weeks) of the programme. The same selection procedures were followed for the second and third cohorts of children participating in the two subsequent years (RG, N=84, N=130 and CG, N=75, N=105), and children were assessed at pre-test and post-test. The main hypotheses of the study were tested using a series of analyses of covariance. The RG was found significantly higher on all the assessments of the OS and on reading age at T2 and T3. However, the effect of the intervention decreased with time, so that at T4 the two groups had similar attainment. Concerning the average group, they had significantly higher reading age than both RG and CG at T4. No main effects of gender, pre-school education or free school meals were detected. Similar short-term effects favouring the RG were demonstrated for the second and third year cohorts. With regard to the study of perceptions, semi-structured interviews were conducted with children (n=93) receiving the intervention, and headteachers (n=51) from all the schools that implemented the programme. An evaluation questionnaire was given to the RR teachers. The analysis of the children's interviews indicated that despite believing that it was difficult, children generally held positive attitudes towards reading and considered being in RR enjoyable and helpful. Furthermore, they viewed themselves principally as good readers. The headteachers' interviews revealed that the effective implementation of RR relied heavily upon the commitment and enthusiasm of the teachers, support from the rest of the school staff and central finding from the government. Finally, the teachers' depiction of the programme was that of a valued professional development programme that helped raise their expectations of pupils and had an additional impact in the classroom.