Poor usability designed of ventilator user interface can easily lead to human error. In this study, we evaluated the usability design of ventilator maintenance user interface and identified problems related to the usability of user interface that could easily cause human error. Sixteen respiratory therapists participated in this usability study. The usability of the ventilator maintenance user interface was evaluated by participants' task performance (task completion time, task error rate), physiological workload (eye-fixation duration) and perceived workload (NASA-TLX), and user experience (questionnaire). For task performance, task completion time and task error rate showed significant differences. For task completion time, significant difference was found when conducting ventilator self-test (p < 0.001), replace the breathing circuit (p = 0.047), and check battery status (p = 0.005). For task error rate, the three ventilators showed significant difference (p = 0.012), and the Serov I showed a significantly higher task error rate than the Boaray 5000D (p = 0.031). For workload, the Serov I was associated with higher physiological and perceived workloads than other ventilators (p < 0.05). For user experience, the Boaray 5000D received better scores among the ventilators in terms of ease to maintain, friendly to maintain, and willingness to use (p < 0.05, respectively). Our study adds available literature for usability evaluation of ventilator maintenance user interface. The results indicate that the maintenance user interface of the Boaray 5000D performed better than the other two tested ventilators. Moreover, the study results also proved that eye-fixation duration can be a reliable tool for evaluating the usability of ventilator user interface.