The oscillatory behaviour of the giant magnetoresistivity versus Cu sublayer thickness was observed in Ni83Fe17/Cu multilayers with an oscillation period of 1 nm independently of buffer layer and substrates used in the experiment. The giant magnetoresistance field sensitivity equal to 0.02%/0e and 0.15%/0e for the first and the second maximum respectively, were found. It was shown that the giant magnetoresistance effect in our multilayers is not only due to the antiferromagnetic coupling between adjacent permalloy layers but also due to the superparamagnetic-like behaviour related to the existence of the intermixed interfacial areas. The giant magnetoresistance (GMR) effect attracts much attention not only from physics but also from practical application. Magnetic metallic multilayers (Mis) with enhanced magnetoresistance form potential new candidates for magnetoresistance heads for magnetic recording storage devices. Such GMR heads require large changes of resistance at low fields for temperatures at above room temperature. Thus, of particular application interest are Mls containing magnetically soft permalloy ((Ni83Fe17 = Py) layers.The present paper describes the systematic study of GMR ratio and saturation fields (Hs) of Py/Cu Mls on Cu and Py sublayer thicknesses (denoted by dcu and dpy , respectively) and on the number of bilayers stack. The B + [Py/Cu]•N multilayers, where B and N denote respectively a buffer layer and a number of Py/Cu bilayers, were prepared by double facing target sputtering method. The Cu and Py seed layers were used as B, and N was equal to 25, 30, 40, and 100. The Mls were deposited onto Si (100) wafer and glass substrate. GMR effect was measured at room temperature in a magnetic field up to 25 kOe. In this paper the GMR ratio is defined as GMR = [R(H = 0) -R(H = Hmax)]/R(H = Hmax)• Figure 1 represents the GMR dependence on Cu sublayer thickness. The figure collects the GMR data for different dp y (0.8 nm < dpy < 1.5 nm), B and N. One can observe an oscillatory behaviour with a period of about 1 nm with two well distinguished maxima. Similar oscillations in GMR have been reported by Lenczowski [1] in Ni80Fe20/Cu Mis, however only for B = 20 nm Cu. From Fig. 1 it is evident that the oscillation period depends neither on N nor B and any kind of substrate used in the present experiment. However, the GMR amplitude was found to depend strongly on dpy .(273)