Background <p>Delayed diagnosis and treatment of type 2 diabetes increases diabetes-related complications, making the high prevalence of undiagnosed type 2 diabetes in Hong Kong an important concern. Point-of-care capillary HbA1c (POC-cHbA1c) testing holds promise as a comparably accurate, convenient, and timely alternative to venous HbA1c (vHbA1c) for type 2 diabetes screening, yet randomized trials are lacking. This study compared the effectiveness of a 2-step active opportunistic screening strategy using POC-cHbA1c versus usual practice employing vHbA1c and multiple clinic visits in detecting type 2 diabetes among at-risk primary care patients. The primary outcomes were to identify the difference in the proportion of type 2 diabetes detected between intervention (POC-cHbA1c) and control (vHbA1c) groups and the uptake rate of POC-cHbA1c versus vHbA1c testing among consenting participants.</p> Methods <p>A cluster randomized controlled trial was conducted in 8 General Out-Patient Clinics between June 2022 and January 2024 using 2-step active opportunistic screening.</p> <p>In step 1, risk factor count, 852 at-risk patients were identified through consecutive sampling during their primary care consultation by specific inclusion and exclusion criteria. In step 2, these at-risk patients then underwent POC-cHbA1c (intervention) or vHbA1c (control) testing. If preliminary HbA1c was ≥ 5.6%, a confirmatory oral glucose tolerance test was offered. Randomization occurred at the clinic level using a random allocation sequence generated by statistical software. Multilevel logistic regression analyses were employed to evaluate the effect of the intervention on the uptake rate, adjusting for patient characteristics and clinic clustering.</p> Results <p>POC-cHbA1c had a higher uptake rate than vHbA1c (76.0% vs 37.5%; OR = 7.06, 95% CI [2.47–20.18], <i>p</i> &lt; 0.001). A greater proportion of type 2 diabetes (4.2% vs 1.4%; <i>p</i> = 0.016) and pre-diabetes (11.8% vs 6.9%; <i>p</i> = 0.015) were detected using POC-cHbA1c versus vHbA1c. POC-cHbA1c was more likely to detect type 2 diabetes/pre-diabetes combined (OR = 1.99, 95% CI [1.01–3.95], <i>p</i> = 0.048). The number-needed-to-screen to detect one additional type 2 diabetes patient with POC-cHbA1c was 61 versus vHbA1c.</p> Conclusions <p>POC-cHbA1c testing was associated with a higher uptake rate and detection of type 2 diabetes versus vHbA1c, underscoring its potential as an effective type 2 diabetes screening strategy in primary care.</p> Trial registration <p>NCT06382363 (retrospectively registered: 2024–04-19).</p>