Plasmodium knowlesi is the major cause of zoonotic malaria in Southeast Asia. Rapid and accurate diagnosis enables effective clinical management. A novel malaria diagnostic tool, Gazelle (Hemex Health, USA) detects haemozoin, a by-product of haem metabolism found in all Plasmodium infections. A pilot phase refined the Gazelle haemozoin identification algorithm, with the algorithm then tested against reference PCR in a larger cohort of patients with P. knowlesi mono-infections and febrile malaria-negative controls. Limit-of-detection analysis was conducted on a subset of P. knowlesi samples serially diluted with non-infected whole blood. The pilot phase of 40 P. knowlesi samples demonstrated 92.5% test sensitivity. P. knowlesi-infected patients (n = 203) and febrile controls (n = 44) were subsequently enrolled. Sensitivity and specificity of the Gazelle against reference PCR were 94.6% (95% CI 90.5–97.3%) and 100% (95% CI 92.0–100%) respectively. Positive and negative predictive values were 100% and 98.8%, respectively. In those tested before antimalarial treatment (n = 143), test sensitivity was 96.5% (95% CI 92.0–98.9%). Sensitivity for samples with ≤ 200 parasites/µL (n = 26) was 84.6% (95% CI 65.1–95.6%), with the lowest parasitaemia detected at 18/µL. Limit-of-detection (n = 20) was 33 parasites/µL (95% CI 16–65%). The Gazelle device has the potential for rapid, sensitive detection of P. knowlesi infections in endemic areas.