In this text, we review some recent developments in econometrics that may be of interest to specialists in other areas than economics, and we discuss some relations of econometric methodology with general themes of scientific methodology and philosophy of science, such as the falsifiability criterion (Popper), the underdetermination of theories by data (Quine) and instrument-alism. In particular, we underscore the contrast between two styles of modelling the parsimonious approach and the statistico-descriptive approach and we analyze the links between statistical testing theory and philosophy of science.