INRiM develops references for atmospheric pollutants (such as greenhouse gases and their precursors) from natural and anthropogenic sources, by using complementary primary methods, like gravimetry and dynamic dilution. It is also active in the field of measurements for seawater quality. It develops new statistical approaches to evaluate conformity and assure traceability for chemical/environmental measurements. It cooperates with other research institutions for the development of measurement methods and the evaluation of uncertainties.

Development of statistical models, uncertainty evaluation, as well as studies on Machine Learning algorithms for metrological applications are activities conducted within the framework of MATHMET, the European Metrology Network for Mathematics and Statistics of which INRIM is a member.