In this study, we randomly assigned struggling readers in special needs education (n = 23; 8-12 years of age) to a dog-assisted reading intervention or a similar intervention without a dog present. Students participated in 30-minute reading sessions twice a week for a total of six weeks. Using two standardised tests we measured reading skills before, immediately after, and three months after the intervention. In addition, the task-related behaviours and emotional expressions of a sub-group of these students (n = 17) were observed during the sessions. Results show that students who received the dog-assisted intervention had a higher increase between the pre- and post-test in their reading scores on two standardised tests, for both single-word reading and full-text reading. Differences between the groups at the follow-up measurement were present, but not significant. Observations indicated that the group receiving the dog-assisted reading intervention showed a longer duration of on-task behaviour and positive emotions during the sessions. No differences in the duration of off-task behaviour and negative emotions were found.