The reliability of grinding performance was assessed in 18 current Emirates Team New Zealand America's Cup sailors in two test sessions separated by 5 h using a custom-built ergometer. Sixteen different grinding conditions that varied by load (Light 39 N·m, Moderate 48 N·m, Heavy 68 N·m), deck heel (Flat 0° control, Downhill 25°, Uphill 25°, Right 25°, Left 25°), and grinding direction (forwards, backwards) were assessed using peak power and external work over 5 s during maximal-effort 8-s grinds. Reliability statistics included the difference in mean (Mdiff), standard error of measurement (SEM), and intraclass correlation coefficients (ICC). External work (SEM = 1.6–6.9%; ICC = 0.91–0.99) was a more consistent performance measure than peak power (SEM = 1.3–9.6%; ICC = 0.84–0.99) across all test conditions. Testing under different load conditions resulted in external work SEMs of 1.6–3.9% with performance more reliable in lighter load conditions. Grinding performance during different heel conditions was less reliable (external work SEMs = 4.6–6.9%). Grinding direction (forward or backward) did not appear to affect performance reliability, although external work was 10–15% higher in forward grinding. Reliability is acceptable across various loads, but testing under different heel conditions may need some protocol development to allow the detection of smaller differences in performance.