Background: A combination of limited access to eye services and low numbers of eye care providers in low- and middle-income (LMIC) populations results in high prevalence of avoidable vision impairment. Poverty, lack of awareness and greater distance are key barriers to accessing services. We investigated the effectiveness of the Peek Community Eye Health system (Peek CEH), a smartphone-based referral system comprising decision support algorithms (Peek Community Screening app), Short Message Service (SMS) reminders and real-time system level reporting. Methods: We conducted a single masked, cluster randomised controlled trial. All participants in a ‘community unit’ defined as a dispensary or health centre with its catchment population were eligible. Community units were randomly allocated (1:1) to receive either the Peek CEH and referral (intervention group) or the current standard with periodic health centre-based outreach clinics and onward referral (control group). Participants in the intervention group were assessed in their houses by screeners and those referred were asked to present for triage assessment in a central location. They also received regular SMS reminders. In both groups, community sensitisation was done followed by a triage clinic at the link health facility four weeks from sensitisation. During triage, participants in both groups were assessed and treated and, if necessary, were given a referral to Kitale Eye Unit, participants in the intervention arm received further SMS reminders. The primary outcome was the number of people per 10,000 population (rate) with eye conditions attending triage at four weeks of sensitization. Primary analysis was by intention to treat, with the intervention effect estimated using a t-test performed on cluster-level rates. Findings: Trial recruitment occurred between November 26, 2018, and June 7, 2019. We randomly selected 36 out of the 66 eligible Community Units; these were then randomised to either the intervention group (68,348 participants) or the control group (60,243 participants). 9,387 (13·7%) participants from the intervention group and 3,070 (5·1%) from the control group attended triage assessment. The mean attendance rate by participants with eye problems was 1,429 per 10,000 in the intervention arm and 522 per 10,000 in the control group, giving an estimated rate difference of 906 per 10,000 (95%CI: 689-1,124; p <0·0001). The mean hospital attendance was 82 per 10,000 in the intervention group and 33 per 10,000 in the control group, estimated rate difference of 49 per 10,000 (95%CI:25-73; p=0·0002). Hospital utilization (secondary care) by the catchment population remained consistent with normal annual levels (80 per 10,000), however a major improvement in the proportion of appropriate utilization was seen, with 13% (baseline 61%) having primary eye care conditions (most managed at triage) and 56% (baseline 8%) having priority vision impairing eye conditions. Interpretation: The Peek Community Eye Health system increased primary care and hospital attendance by people with eye problems compared with the standard approach in communities. This indicates the potential of this mHealth package to improve uptake of eye services and guide appropriate task sharing, through better matching of need to services. Trial Registration: This trial is registered with Pan African Clinical Trial Registry, number PACTR 201807329096632. Funding Statement: This study was funded through the Commonwealth Eye Health Consortium (CEHC), with funding from the Queen Elizabeth Diamond Jubilee Trust’s “Avoidable Blindness” programme. The CEHC is administered through the International Centre for Eye Health (ICEH) based at the London School of Hygiene & Tropical Medicine. MJB is supported by the Wellcome Trust 207472/Z/17/Z. Declaration of Interests: The Peek Vision Foundation (09919543) is a registered charity in England and Wales (1165960) with a wholly owned trading subsidiary, Peek Vision Ltd (09937174). Professor Matthew Burton is a Trustee of The Peek Vision Foundation and Dr Andrew Bastawrous is CEO of The Peek Vision Foundation and Peek Vision Ltd. HR is an advisor to Peek Vision Ltd. CB works at Peek Vision Ltd. All other authors declare no conflict of interest. This submission has not been published anywhere previously and is not simultaneously being considered for any other publication. Ethics Approval Statement: The study was approved by the Moi University Institutional Research and Ethics Committee, Kenya and the London School of Hygiene & Tropical Medicine Ethics Committee, UK. Permission was also granted by department of Health, Trans Nzoia County, Kenya. The study adhered to the principles of the Declaration of Helsinki on Ethics. Written informed consent from all participants was obtained by a trained field research team before enrolment; parents or guardians consented for children.