Dutch children at higher familial risk of reading disability received a home-based intervention programme before formal reading instruction started to investigate whether this would reduce the risk of dyslexia. The experimental group (n=23) received a specific training in phoneme awareness and letter knowledge. A control group (n=25) received a non-specific training in morphology, syntax, and vocabulary. Both interventions were designed to take 10 min a day, 5 days a week for 10 weeks. Most parents were sufficiently able to work with the programme properly. At post-test the experimental group had gained more on phoneme awareness than the control group. The control group gained more on one of the morphology measures. On average, these specific training results did not lead to significant group differences in first-grade reading and spelling measures. However, fewer experimental children scored below 10th percentile on word recognition. Copyright © 2008 John Wiley & Sons, Ltd.