Background and aims: The incidence of total knee arthroplasty (TKA) is increasing, especially among younger working-age patients. However, dissatisfaction rates in this population are higher than among older patients. The aim of this study was to assess the rates of dissatisfaction and persistent pain after TKA and to evaluate those factors that predict these outcomes. Material and methods: In total, 186 patients undergoing unilateral TKA aged 65 years or less were enrolled into this prospective observational study with 2-year follow-up. To assess the outcome, the visual analogue scales regarding satisfaction and persistent pain at rest and during exercise were used. In addition, the association between patients´ demographics, radiographic severity of knee osteoarthritis (OA), patient-reported outcome measures (PROMs) and dissatisfaction and persistent pain were tested by univariate logistic regression analysis. Mild OA was defined as Kellgren-Lawrence (KL) grade 2 and severe OA as KL grade 3–4. Furthermore, multiple logistic regression analysis was also conducted to test statistically significant relations. Results: After 2 years, 12 % (n = 23) of patients were dissatisfied with the outcome of TKA, 27 % (n = 50) reported persistent pain during exercise and 10 % (n = 18) at rest. Patients with mild knee OA were significantly more dissatisfied (28.6 %) than patients with more severe OA (8.7 %) (p = 0.003). Younger patients had an increased risk for both dissatisfaction and persistent pain. Apart from KOOS Quality of Life, poor preoperative KOOS subscores were also predictive for these outcomes. Conclusion: Mild radiographic knee OA was the main predicting factor for dissatisfaction after TKA. Thus, performing TKA for such patients should be carefully considered. Furthermore, these patients should be informed about the increased risk for dissatisfaction and the same seems to apply to younger patients. Interestingly, when TKA is performed for patients with more severe knee OA, the satisfaction rates seem to be somewhat higher than those previously reported. Trial registration: The study was retrospectively registered with ClinicalTrials.gov (registration number NCT03233620) on 28 July 2017.