Mouse dynamics, the analysis of individuals' distinctive mouse actions, has been proposed as a biometric to discriminate legitimate users from impostors. Some researchers have explored this domain and showed promising results, but few examined the effectiveness and applicability of this technique, particularly regarding the inconsistent evaluation conditions, unacceptably long authentication time, and unshared data set. In this work, we perform a benchmark study of the baseline capabilities of mouse dynamics for static authentication. Using the data collected under a tightly-controlled environment from 26 subjects, we develop a repeatable evaluation procedure to investigate the performance of this technique at a continuum of authentication times and various classifiers. The results show that the longer the authentication time, the better the accuracy. The equal-error rate reduces from 14.26% to 2.64%, as the authentication time increases from 11 seconds to 110 seconds. Further, the technique is able to meet the European standard for commercial biometric technology if a longer authentication time is allowed. We also publish our data set and discuss a number of avenues for future research, which we believe are necessary to advance the state of the art in this area.