The global burden of MVC injuries and deaths among vulnerable road users, has led to the implementation of prevention programmes and policies at the local and national level. MVC epidemiological research is key to quantifying MVC burden, identifying risk factors and evaluating interventions. There are, however, several methodological considerations in MVC epidemiological research. This manuscript collates and describes methodological considerations in MVC epidemiological research, using examples drawn from published studies, with a focus on the vulnerable road user population of children and adolescents. Methodological considerations in MVC epidemiological research include the availability and quality of data to measure counts and calculate event rates and challenges in evaluation related to study design, measurement and statistical analysis. Recommendations include innovative data collection (eg, naturalistic design, stepped-wedge clinical trials), combining data sources for a more comprehensive representation of collision events, and the use of machine learning/artificial intelligence for large data sets. MVC epidemiological research can be challenging at all levels: data capture and quality, study design, measurement and analysis. Addressing these challenges using innovative data collection and analysis methods is required.