In this paper, we propose an asynchronous paradigm for controlling a car using steady-state visual evoked potential (SSVEP)-based brain-computer interface (BCI) and conduct experimental tests on real car outside the laboratory. The paradigm uses six stimulation frequencies to classify targets by canonical correlation analysis (CCA) method and generates multi-task vehicle control strategies, including left and right turn signals, wipers, horns, doors and hazard lights. Four healthy volunteers participated in the online car control experiment, and the average correct rate reached 88.43%. Subject S1 showed the most satisfactory BCI-based performance, and its true positive rate and false positive rate were in line with expectations. The research shows the feasibility and effectiveness of the paradigm in automotive control applications, which lays the foundation for future research and development of related brain-controlled automotive technologies, thereby helping individuals with mobility impairments to provide supplements or alternatives, and can also provide an auxiliary vehicle driving strategy for healthy people.