Accessing a database using natural language has the potential to broaden information retrieval, making it accessible to users without SQL knowledge. This task, known as text-to-SQL, can also benefit the area of process mining, which provides tools to extract valuable insights from event logs. However, the text-to-SQL task in the process mining domain has not been fully explored. In this paper, we evaluate the text-to-SQL task using the text \(_2\) SQL \(_4\) PMdataset, a process mining domain-specific dataset built to serve as a benchmark for text-to-SQL implementations on process mining domain. We evaluated three large language models using different prompt strategies and representations. A detailed analysis of the results was conducted, providing insights for understanding the usability and feasibility of applying text-to-SQL on process mining domain.