Fairness and justice are key aspects in the evaluation of language assessments. In this paper, I argue that fairness and justice are subjective perceptions of test takers. Consistent with Kunnan (2018), fairness was conceived as a test quality, while justice was a quality of the social entity administering the test; which in this study was the language program. Test fairness was conceptualized as a multidimensional construct—distributive fairness (how fairly learners perceive their score relative to their effort) procedural fairness (equality of procedures administered during the assessment), interactional fairness (respectful interpersonal treatment of the students by the test administrator), and informational fairness (how fairly information about the assessment and its procedures were provided). In total, 83 university L2 learners in Taiwan completed an online questionnaire gauging perceptions of a single test administration. Regression analysis revealed that procedural and interactional fairness were predictive of justice perceptions of the language program. Distributive fairness shared a relationship with justice perceptions but was not predictive. These findings suggest that participants made justice judgments of their language programs based on how fairly the procedures used to administer a single test were carried out and how respectfully the instructor administering the test treated them.