To evaluate whether C-reactive protein (CRP), including variation within the normal range, is predictive of long-term disease outcome in Juvenile Idiopathic Arthritis (JIA). Consecutive patients with newly diagnosed JIA were included prospectively from defined geographic areas of the Nordic countries from 1997 to 2000. Inclusion criteria were availability of a baseline serum sample within 12 months after disease onset and 8-year clinical assessment data. Systemic onset JIA was not included. CRP was measured by high-sensitive ELISA (detection limit of 0.2 mg/l). One hundred and thirty participants with a median follow-up time of 97 months (range 95–100) were included. At follow-up, 38% of the patients were in remission off medication. Absence of remission was associated with elevated level of CRP at baseline (odds ratio (OR) 1.33, confidence interval (CI) 1.08–1.63, p = 0.007). By applying a cutoff at the normal upper limit (>10 mg/l), the risk of not achieving remission was increased to an OR of 8.60 (CI 2.98–24.81, p < 0.001). Variations of CRP within the normal range had no predictive impact on disease activity at follow-up. Baseline levels of ESR were available in 80 patients (61%) and elevated ESR was associated with absence of remission in a multivariable logistic regression analysis (OR 2.32, CI 1.35-4.00, p = 0.002). This results of this study indicate that baseline CRP concentrations above 10 mg/l are predictive of a poor outcome at 8-year follow-up. We could not demonstrate any predictive value of CRP variations within the normal range.