BACKGROUND: GET.ON (HelloBetter) treatment interventions have been shown to be efficacious in multiple randomized controlled trials. OBJECTIVE: This study evaluated the effectiveness of 2 GET.ON interventions, GET.ON Mood Enhancer and GET.ON Stress, in a national digital mental health service implemented across Germany. METHODS: Following an initial web-based questionnaire, participants were allocated to either intervention based on their baseline symptom severity and personal choice and received a semistandardized guided, feedback-on-demand guided, or self-guided version of the treatment. Uncontrolled routine care data from 851 participants were analyzed using a pretest-posttest design. Half of the participants (461/851, 54.2%) were allocated to the stress intervention (189/461, 41% semistandardized; 240/461, 52% feedback on demand; and 32/461, 6.9% self-guided), and almost all participants in the mood intervention (349/352, 99.2%) received semistandardized guidance. RESULTS: Results on depression-related symptom severity indicated a reduction in reported symptoms, with a large effect size of d=-0.92 (95% CI -1.21 to -0.63). Results on perceived stress and insomnia indicated a reduction in symptom severity, with large effect sizes of d=1.02 (95% CI -1.46 to -0.58) and d=-0.75 (95% CI -1.10 to -0.40), respectively. A small percentage of participants experienced deterioration in depression-related symptoms (11/289, 3.8%), perceived stress (6/296, 2%), and insomnia (5/252, 2%). After completing treatment, 51.9% (150/289) of participants showed a clinically reliable change in depression-related symptoms, whereas 20.4% (59/289) achieved a close to symptom-free status. Similar improvements were observed in perceived stress and insomnia severity. Guidance moderated the effectiveness of and adherence to the interventions in reducing depressive symptom severity. Effect sizes on depression-related symptom severity were d=-1.20 (95% CI -1.45 to -0.93) for the semistandardized group, d=-0.36 (95% CI -0.68 to -0.04) for the feedback-on-demand group, and d=-0.83 (95% CI -1.03 to -0.63) for the self-guided group. Furthermore, 47.6% (405/851) of the participants completed all modules of the intervention. Participant satisfaction was high across all patient groups and both interventions; 89.3% (242/271) of participants would recommend it to a friend in need of similar help. Limitations include the assignment to treatments and guidance formats based on symptom severity. Furthermore, part of the differences in symptom change between groups must be assumed to be due to this baseline difference in the measures. CONCLUSIONS: Future digital health implementation and routine care research should focus on monitoring symptom deterioration and other negative effects, as well as possible predictors of deterioration and the investigation of individual patient trajectories. In conclusion, this study supports the effectiveness of tailored digital mental health services in routine care for depression- and stress-related symptoms in Germany. The results highlight the importance of guidance in delivering internet-based cognitive behavioral therapy interventions and provide further evidence for its potential delivered as web-only solutions for increasing access to and use of psychological treatments.