The aim of this study was to investigate the effects of embodied learning on children's pre-reading and word reading skills. We conducted a three-armed randomized controlled trial including two intervention groups and one control group. One hundred forty-nine children from grade 0 (5-6 years old) who had just started school were recruited from 10 different classes from four elementary schools. Within each class, children were randomly assigned to receive teaching of letter-sound couplings and word decoding either with whole-body movements (WM), hand movements (HM), or no movements (CON) over an 8-week period. Children were evaluated on pre-reading, word reading, and motor skills before (T1), immediately after (T2), and after 17-22 weeks of retention period (T3) following the intervention. Between-group analysis showed a significant improvement in children's ability to name letter-sounds correctly from T1 to T2 (p < 0.001) and from T1 to T3 (p < 0.05) for WM compared to CON. HM and WM improved significantly in naming conditional letter-sounds from T1 to T2 (p < 0.01, p < 0.01) compared to CON and from T1 to T3 for the HM group compared to CON (p < 0.05). We did not find an effect on word reading or a correlation between motor skill performance and reading. Results from the present study suggest that there are beneficial effects of using whole-body movements for children. Hand motor movements indeed also had a performance effect on letter-sound knowledge; however, the whole-body movements had longer-lasting effects. We do not see an effect on whole word reading.