The Nuffield Early Language Intervention is designed to improve the spoken language ability of children during the transition from nursery to primary school. It is targeted at children with relatively poor spoken language skills. Three sessions per week are delivered to groups of two to four children starting in the final term of nursery and continuing in the first two terms of reception in primary school. Children in primary school also attend an additional two 15-minute individual sessions per week. All sessions focus on listening, narrative and vocabulary skills. Work on phonological awareness is introduced in the final ten weeks. The intervention was developed by researchers from the University of York with funding from the Nuffield Foundation. The communications charity I CAN was enlisted to train teaching assistants and nursery staff to deliver the programme. This report evaluates the I CAN-led model for the 30-week programme described above and also a shorter 20-week version delivered only in reception year. The impact of these two programmes on the language skills of 350 children in 34 schools was tested using a randomised controlled trial design. Schools with attached nursery schools or nursery classes in Yorkshire and the South East were recruited to the trial in 2013. Children identified as having relatively low language skills were randomly allocated to the 30-week programme, the 20-week programme or standard provision. The qualitative fieldwork carried out as part of the project involved interviews with a total of 12 staff in 8 of the 34 participating schools. Key conclusions include: (1) The Nuffield Early Language Intervention had a positive impact on the language skills of children in the trial. This is true for both the more expensive, 30-week version, starting in nursery, and the 20-week version, delivered only in school; (2) Children receiving the 30-week version experienced the equivalent of about four months of additional progress, compared with about 2 months additional progress for the 20-week version. Both results are unlikely to have occurred by chance, though results for the 30-week version are more secure; (3) The evaluation did not provide reliable evidence that either version of the programme had a positive impact on children's word-level literacy skills; (4) Teaching assistants delivering the programme reported that they found it difficult to devote enough time to it, and that support from senior staff was required to protect the programme time; and (5) Staff in participating schools reported that the programme had a positive impact on children's language skills and confidence. They thought that the factors which contributed to this included the small-group format, the activities covered, and the focus on narrative and vocabulary work.