During the course of a hypothetical severe Pressurized Water Reactor accident, pressurization of the containment occurs and hydrogen can be produced by the reactor core oxidation. Inner spray systems are used in order to reduce the pressure, mix the gaseous mixture and collect the fission products on the droplets. Several phenomena are involved during the spray injection, making the modelling difficult (for lumped-parameter codes as well as for CFD codes) and the experiments or the experimental facility difficult to scale without distortion. The French Institute for Radiprotection and Nuclear Safety (IRSN) is conducted experimental and numerical separate-effect studies to understand the different phenomena involved during the course of an accident . For this purpose, a facility (TOSQAN, 7m3, 1.5m i.d., 4.8m height) has been designed to reproduce typical thermal-hydraulic conditions. Spray interaction with the atmosphere is studied in the TOSQAN vessel, neglecting water interaction with walls and compartments (separate-effect test). A scaling study has thus been performed in the frame of a European Concerted Action (SCACEX, ). The objective of this paper is to propose scaling rules for spray studies in large enclosures such as TOSQAN.