We propose a naming convention for research data files and folders, intended as a practical tool to achieve reasonable quality without using too much time designing one’s own. It is by no means mandatory and may be adapted freely if needed. Regardless of the present suggestion, it is advisable for all collaborators of a project or research group to follow a consistent practice. This convention is intended for experimental, observational or computed data. We present how to build the full path (parent folders + file name) as a sequence of simple information elements, a few of which can be re-ordered if necessary for the project. We then demonstrate how to use the convention in practice though plausible fictitious examples from various research fields. The convention is not directly applicable for source code and similar files, which usually follow their own rules and practices. However, it is possible to use it for parts of a dataset, see example 4 for a case that combines code and data files.