In order to investigate the system availability of the human-computer interaction interface for elderly drivers during driving, this study employs questionnaires and in-depth interviews to identify common interactive tasks performed by elderly drivers while driving. Experimental models were selected, and elderly drivers were enlisted to partake in real vehicle tests. Data on task completion times during human-computer interactions were collected, and a multiple linear regression model was constructed by integrating the SUS scale. This model was used to analyze the impact of each independent variable’s task completion time on the availability of the dependent variable system. Based on this foundation, additional screen samples were expanded for verification. Control performance across four typical vehicle screens was horizontally compared, and causal analysis was conducted using eye movement data. The findings reveal that the completion time of each human-computer interaction task for elderly drivers during driving significantly affects system availability, with the exception of air conditioning settings. Notably, the completion time of the navigation setting task exhibits the greatest impact on system availability, followed by broadcast radio. Observational speed and call-related tasks have a slightly smaller impact. The human-computer interaction system in vehicles equipped with dual-screens demonstrates the highest availability, suggesting its suitability for elderly drivers.