BACKGROUND: Patellofemoral pain syndrome (PFP) is a common musculoskeletal condition that has a tendency to become chronic and problematic in a proportion of affected individuals. The objective of this study was to identify prognostic factors that may have clinical utility in predicting poor outcome on measures of pain and function in individuals with PFP. METHODS: A prospective follow-up study was conducted of 179 participants in a randomised clinical trial. Nine baseline factors (age, gender, body mass index, arch height, duration of knee pain, worst pain visual analogue scale, Kujala Patellofemoral Score (KPS), functional index questionnaire (FIQ), step down repetitions) were investigated for their prognostic ability on outcome assessed at six, 12 and 52 weeks (worst pain, KPS and FIQ). Factors with significant univariate associations were entered into multivariate linear regression models to identify a group of factors independently associated with poor outcome. RESULTS: Long symptom duration was the most consistent predictor of poor outcome over 52 weeks rated on the KPS and the FIQ (beta-0.07, 95% confidence interval -0.1 to -0.03, p < 0.000; and -0.02, -0.03 to -0.01, p < 0.000, respectively). Worse KPS at baseline was predictive of outcome at six, 12 and 52 weeks. Gender, body mass index and arch height were generally not associated with outcome (univariate analysis), while age, worst pain, FIQ and step downs were excluded during multivariate analyses. CONCLUSIONS: Patients presenting with PFP of long duration who score worse on the KPS have a poorer prognosis, irrespective of age, gender and morphometry. These results suggest that strategies aimed at preventing chronicity of more severe PFP may optimise prognosis.