Seals are extremely useful devices to prevent fluid leakage. We present experimental results for the leak-rate of rubber seals, and compare the results to a novel theory, which is based on percolation theory and a recently developed contact mechanics theory. We find good agreement between theory and experiment.A seal is a device for closing a gap or making a joint fluid tight [1]. Seals play a crucial role in many modern engineering devices, and the failure of seals may result in catastrophic events, such as the Challenger disaster. In spite of its apparent simplicity, it is not easy to predict the leak-rate and (for dynamic seals) the friction forces [2] for seals. The main problem is the influence of surface roughness on the contact mechanics at the seal-substrate interface. Most surfaces of engineering interest have surface roughness on a wide range of length scales [3], e.g, from cm to nm, which will influence the leak rate and friction of seals, and accounting for the whole range of surface roughness is impossible using standard numerical methods, such as the Finite Element Method.In this paper we present experimental results for the leak-rate of rubber seals, and compare the results to a novel theory [3,4,5], which is based on percolation theory and a recently developed contact mechanics theory [6,7,8,9,10,11,12], which accurately takes into account the elastic coupling between the contact regions in the nominal rubber-substrate contact area. Earlier contact mechanics models, such as the GreenwoodWilliamson[13] model or the model of Bush et al [14], neglect this elastic coupling, which results in highly incorrect results [15,16], in particular for the relations between the squeezing pressure and the interfacial separation [17]. We assume that purely elastic deformation occurs in the solids, which is the case for rubber seals.Consider the fluid leakage through a rubber seal, from a high fluid pressure P a region, to a low fluid pressure P b region, as in Fig. 1. Assume that the nominal contact region between the rubber and the hard countersurface is rectangular with area L x × L y . We assume that the high pressure fluid region is for x < 0 and the low pressure region for x > L x . We "divide" the contact region into squares with the side L x = L and the area A 0 = L 2 (this assumes that N = L y /L x is an integer, but this restriction does not affect the final result). Now, let us study the contact between the two solids within one of the squares as we change the magnification ζ. We define ζ = L/λ, where λ is the resolution. We study how the