© 2015, Australasian Medical Publishing Co. Ltd. All rights reserved. Objectives: To determine whether a combination of point-of-care (POC) and laboratory glycated haemoglobin A (HbA1c) testing (HbA1c algorithm) is more effective in testing for diabetes in everyday practice in remote Australian Aboriginal primary health care, by providing a more rapid definitive result and diagnosing more cases than the standard glucose algorithm. Design: Cross-sectional study that independently classified participants using both diagnostic algorithms and compared their outcomes. Participants: Two hundred and fifty-five Aboriginal Australians aged 15 years or more without confirmed diabetes and due for diabetes testing at participating clinics. Setting: Six primary health care sites in the Kimberley region of Western Australia from 1 September 2011 to 30 November 2013. Main outcome measures: Number of participants with a definitive test result, a completed algorithm and a diagnosis of diabetes; time taken to deliver a test result. Results: Participants were significantly more likely to have a definitive result within 7 days (249 v 199 of 255 participants; P <0.001), be followed up if an initial laboratory measurement was abnormal (92 v 74 of 167 participants; P = 0.005), and be diagnosed with diabetes (15 v 4 of 255 participants; P = 0.003) using the HbA1c than with the glucose algorithm. Eight participants subsequently diagnosed with diabetes (four using the HbA1c test, four with additional oral glucose tolerance tests that would not normally have been requested) were incorrectly classified as normal by the glucose algorithm. No participants with normal HbA1c measurements were subsequently diagnosed with diabetes. Conclusions: Use of POC HbA1c testing and collection of venous blood on the same day for a confirmatory laboratory HbA1c testing if the POC HbA1c value is abnormal may simplify diabetes testing in remote areas, provide more timely diagnoses, and increase case detection.