We independently assign a non-negative value, as a capacity for the quantity of flows per unit time, with a distribution F to each edge on the Z d lattice. We consider the maximum flows through the edges from a source to a sink, in a large cube. In this paper, we show that the ratio of the maximum flow and the size of source is asymptotic to a constant. This constant is denoted by the flow constant.AMS classification: 60K 35. Key words and phrases: maximum flow and minimum cut, random surfaces, cluster boundary, and first passage percolation. A(ω) or N(ω) for each, respectively. Let P be the corresponding product measure on Ω. The expectation with respect to P is denoted by E(·). For simplicity, we assume that τ (e) has a short tailfor some η > 0. For each finite graph B with vertices and edges, we may think of τ (e) as the non-negative capacity for the quantity of fluid that may flow along e ∈ B in unit time, where an edge in a set means that the two vertices of the edge belong to the set. Let S and T be two disjoint sets in B, called the source and the sink. A flow (see Kesten (1987); Grimmett (1999)), from a vertex set S to another vertex set T in B, is an assignment of a non-negative number f (e) and an orientation to each edge e = (v, w) of B such thatsatisfies I(v) = 0 for all vertices v ∈ S ∪ T, where the first summation (with respect to the second summation) is calculated over all neighbors w of v, which e(v, w) is oriented away from (respectively toward) v. Thus fluid is conserved at all vertices except, possibly, at sources and sinks. In other words, the current flowing into a vertex v ∈ S ∪ T must equal to the current flowing out. This basic assumption is called Kirchhoff's law in physics. A flow is admissible if f (e) ≤ τ (e) for all edges e,