Objective: To determine and compare the proportion of patients treated in a primary care setting that have knee osteoarthritis (OA) according to the European League Against Rheumatism (EULAR), the American College of Rheumatology (ACR), and the National Institute for Health and Care Excellence (NICE) clinical classification criteria. Design: A cross-sectional analysis of baseline data from 13,459 patients with knee symptoms or functional limitations associated with OA participating in the Good Life with osteoArthritis in Denmark (GLA:D®) treatment program. The prevalence of knee OA according to the EULAR, ACR and NICE criteria were calculated in all participants and in the subgroup of patients with self-reported radiographic changes associated with knee OA (n = 10,651 or 79%). Results: Mean age (SD) was 65.3 (9.8) years, median (IQR) duration of symptoms 12 (6–36) months and mean pain intensity (0–100; SD) 46.5 (22.1) mm. 39% fulfilled all three sets of criteria. 48%, 52% and 89% fulfilled the EULAR, ACR, and NICE criteria for having knee OA, respectively. In the subgroup with self-reported radiographic changes, the corresponding numbers were 49%, 54% and 90%. Conclusions: While the EULAR and ACR criteria only identified around half of those, with or without self-reported radiographic knee OA, that were treated because of symptoms or functional limitations associated with knee OA, the NICE criteria identified most patients. The results indicate that the NICE criteria are relevant and appropriate to identify individuals treated for knee OA in primary care.