In this paper, I describe widely used methods in Scandinavian police research, discuss strengths and weaknesses and suggest methods for future police research. Thick, in-depth descriptions and closeness to the field are strengths in today's police research, whereas vulnerability to the Hawthorne effect, the lack of representativeness and limited possibilities for causal claims are weaknesses. I suggest new methods for improving police research: natural variation designs, vignette studies and field experiments. In addition, extended use of an existing data source, register data, is discussed. Examples are given from police research and connected research fields, including research on other professions. Greater methodological variation can address new questions in future police research and widen the horizons of Scandinavian police research. Especially, new methods and data can improve the possibilities for making causal claim and improve the external validity.