Background: Research shows that internet-based cognitive behavioural therapy (iCBT) is an effective treatment for depression. However, little is known about how the length of the text material in iCBT affects outcomes. Objective: The aim of this study was to test whether a condensed iCBT version for depression would be noninferior to the existing full-text version in reducing depressive symptoms at post-treatment. We also wanted to test non-inferiority for secondary outcomes and explore reading speed and ADHD symptoms as potential moderators. Method: A single-blind randomized controlled trial was conducted (N = 267) comparing two versions of guided iCBT for depression; full-text (around 60,000 words) and condensed (around 30,000 words, with the option to listen to the text). Estimated between-group effect sizes and their confidence intervals for depression, anxiety and quality of life, were compared to a pre-determined non-inferiority margin (ES = 0.4). Moderation analyses of reading speed and ADHD symptoms were conducted. Results: The condensed version of iCBT was non-inferior to the full-text version on post-treatment measures for depressive symptoms (95 % CI = - 0.42-0.24), anxiety symptoms (95 % CI = - 0.24-0.32), and quality of life (95 % CI = - 0.09-0.49). Non-inferiority was inconclusive for depressive symptoms at the one-year follow-up (95 % CI = - 0.60-0.47). There was no significant moderation effects of reading speed (p = 0.06) or ADHD symptoms (p = 0.11) on depressive symptoms. Conclusion: These results indicate that a condensed version of iCBT for depression is as effective at treating depression as the full-text version. By shortening texts, iCBT may be made available to more people. Due to unequal dropout rates between the groups, these results are preliminary and need to be replicated.