Experiments performed with two coupled uranium (93.16 wt% 235U) metal cylinders (17.77-cm o.d., 5.08 cm thick) are the first application to coupled systems of the 252Cf-source-driven neutron noise analysis method for obtaining the subcritical neutron multiplication factor. These coaxial cylinders were separated axially by various thicknesses of either air or borated plaster between the flat surfaces. In all measurements, the 252Cf neutron source was located at the center of the outer flat surface of one cylinder, and the two detectors were located in three configurations: (a) both adjacent to the radial surface of the cylinder with the source, (b) both detectors adjacent to the radial surface of the cylinder without the source, and (c) one detector adjacent to the radial surface of each cylinder. A ratio of spectral densities obtained with the source and detectors adjacent to the cylinder with the source can be interpreted using point kinetics to obtain the subcritical neutron multiplication factor. However, when the source and detectors are placed on different cylinders, a coupled kinetics model is required to interpret the ratio of spectral densities. The cross-power spectral densities between detector and source positioned on different cylinders depend on the neutronic coupling and approach zero as the coupling does. By comparing the subcriticality from the measurements performed with borated plaster separating the uranium cylinders to those separated by air, it was found that the neutron multiplication factor was always increased by the insertion of borated plaster between the cylinders, regardless of their separation.