OBJECTIVE: This study sought to assess the effect of reading augmented reality (AR) books on salivary cortisol levels in hospitalized pediatric patients compared to reading a standard children's book. METHODS: This was a randomized, two-period, cross-over trial in hospitalized children aged 7-11 years. AR books currently in the market were used as intervention. Complete block randomization was used to randomize the order of the intervention. Children allocated to the 'AR-first' group received the book, a tablet and were left to interact independently with the technology for an hour. After a 48 -h wash-out period, children received a standard book. 'Standard-book-first' group received only the standard book and after wash-out received the tablet and the AR book. Salivary cortisol and a validated visual analogue scale (VAS) for psychological stress were assessed at the beginning and at the end of each intervention. RESULTS: A total of 29 children were recruited in the study. One was lost during follow up. Cortisol levels decreased after the AR intervention (P = 0.019). Nevertheless, the decrease was not greater than the one associated to reading the standard book. VAS scores increased after the AR intervention (P < 0.001). DISCUSSION: There is evidence of order and sequence effects that might explain results. First assessment of AR-based interventions on stress. Results justify further research. CONCLUSIONS: There was no evidence that reading AR books diminished cortisol levels more than reading a standard book. AR-books improved VAS score for psychological stress compared to a standard book.