Vitamin K antagonist (VKA) therapy for stroke prevention in atrial fibrillation (AF) requires monitoring of the international normalized ratio (INR). We evaluated the agreement between two INR audit parameters, frequency in range (FIR) and proportion of time in the therapeutic range (TTR), using data from a global population of patients with newly diagnosed non-valvular AF, the Global Anticoagulant Registry in the FIELD-Atrial Fibrillation (GARFIELD-AF). Among 17 168 patients with 1-year follow-up data available at the time of the analysis, 8445 received VKA therapy (±antiplatelet therapy) at enrolment, and of these patients, 5066 with ≥3 INR readings and for whom both FIR and TTR could be calculated were included in the analysis. In total, 70 905 INRs were analysed. At the patient level, TTR showed higher values than FIR (mean, 56·0% vs 49·8%; median, 59·7% vs 50·0%). Although patient-level FIR and TTR values were highly correlated (Pearson correlation coefficient [95% confidence interval; CI], 0·860 [0·852-0·867]), estimates from individuals showed widespread disagreement and variability (Lin's concordance coefficient [95% CI], 0·829 [0·821-0·837]). The difference between FIR and TTR explained 17·4% of the total variability of measurements. These results suggest that FIR and TTR are not equivalent and cannot be used interchangeably.