PURPOSE: No outcome measurements have previously been designed to assess patient-perceived patellar instability. The purposes of this study were to address this limitation and to describe the development and validation of the Norwich Patellar Instability (NPI) scores, a self-administered 19-item questionnaire to assess perceived patellar instability. METHODS: A previous study assessed activities that aggravated symptoms in individuals with patellar dislocation and instability symptoms. These reported activities were ranked in order of severity, and a weighting system was calculated. The NPI score was introduced to routine clinical practice in three institutions. One hundred and two people who had experienced a lateral patellar dislocation completed 102 NPI questionnaires. The completed NPI score was correlated with the Knee injury and Osteoarthritis Outcome Score (KOOS) to evaluate divergent validity and with the J-sign, patellar mobility, apprehension test, Beighton score, Kujala Patellofemoral Disorder Score and Lysholm Knee Score to evaluate convergent validity. Internal consistency of the 19 items was also calculated. RESULTS: The results indicated a moderate correlation between the NPI score and the Kujala Patellofemoral Disorder Score (rho = -0.66; p < 0.01) and Lysholm Knee Score (rho = -0.54; p = 0.03), suggesting good convergent validity. There was a little correlation between the KOOS and NPI score, indicating divergent validity (rho = -0.02 to -0.17; n.s.). There was high internal consistency (Cronbach's alpha = 0.93). CONCLUSION: These results indicate the NPI score to be a valid tool to assess patellar instability for individuals following patellar dislocation. Further study is now required in order to assess the reliability and responsiveness of this new outcome measure.