Curvilinear structures are useful features in a variety of applications. Compared to other commonly used features such as edges, there is relatively few work on curvilinear structure detection and its performance evaluation. In this paper we propose a novel supervised methodology for evaluating the performance of curvilinear structure detection algorithms. We consider the two aspects of performance, namely detection rate and detection accuracy, separately, in contrast to their mixed handling in earlier approaches that typically produces biased impression of detection quality. By doing so, the proposed performance measures give us a more informative and precise performance characterization. We will demonstrate the advantages of our approach using both synthetic and real examples.