BACKGROUND: Depression is the most prevalent mental health condition in older adults. However, not all evidence-based treatments are easily accessible. Web-delivered cognitive behavioral therapy (wCBT) facilitated by laypersons is a viable treatment alternative. OBJECTIVE: This randomized controlled trial aims to evaluate the efficacy of a novel wCBT program, Empower@Home, supported by trained lay coaches, against a waitlist attention control. Empower@Home is among the very few existing wCBT programs specifically designed for older adults. The primary objective was to assess the efficacy of the intervention compared with attention control. The secondary objective was to evaluate the program's impact on secondary psychosocial outcomes and explore potential change mechanisms. METHODS: Older adults (N=70) were recruited via web-based research registries, social media advertisements, and community agency referrals and randomly assigned to either the intervention or control group in a 1:1 allocation ratio. The intervention group received access to Empower@Home, which included 9 web-delivered self-help lessons and weekly telephone coaching sessions by a trained layperson over 10 weeks. The control group received weekly friendly phone calls and depressive symptom monitoring. The primary clinical outcome was the severity of depressive symptoms assessed using the Patient Health Questionnaire-9. The secondary clinical outcomes included anxiety, anger, social isolation, insomnia, pain intensity, and quality of life. Linear mixed modeling was used to determine the treatment effects on depression, and 2-tailed t tests were used to assess within-group changes and between-group differences. RESULTS: Most participants in the intervention group completed all 9 sessions (31/35, 89%). The usability and acceptability ratings were excellent. The intervention group had a large within-group change in depressive symptoms (Cohen d=1.22; P<.001), whereas the attention control group experienced a medium change (Cohen d=0.57; P<.001). The between-group effect size was significant, favoring the intervention group over the control group (Cohen d=0.72; P<.001). In the linear mixed model, the group-by-time interaction was statistically significant (b=-0.68, 95% CI -1.00 to -0.35; P<.001). The treatment effects were mediated by improvements in cognitive behavioral therapy skills acquisition; behavioral activation; and satisfaction with the basic psychological needs of autonomy, competence, and relatedness. Furthermore, the intervention group showed significant within-group improvements in secondary psychosocial outcomes, including anxiety (P=.001), anger (P<.001), social isolation (P=.02), insomnia (P=.007), and pain (P=.03). By contrast, the control group did not experience significant changes in these outcome domains. However, the between-group differences in secondary outcomes were not statistically significant, owing to the small sample size. CONCLUSIONS: Empower@Home, a wCBT program supported by lay coaches, was more efficacious in reducing depressive symptoms than friendly telephone calls and depression symptom monitoring. Future studies should examine the effectiveness of the intervention in community and practice settings using nonclinician staff already present in these real-world settings as coaches. TRIAL REGISTRATION: ClinicalTrials.gov NCT05593276; https://clinicaltrials.gov/ct2/show/NCT05593276. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR2-10.2196/44210.