The tendency for cyclization of nonlinear network molecules is quantitatively expressed by the cyclization probability, defined as the ratio of cyclization to total reaction rate. In contrast to branching, the rate of cyclization depends on the configurational statistical mechanics of segments joining functional groups. Direct integration of the joint configurational probability density in the subspace of encounters yields the distribution of the number of configurationally formed intramolecular functional group encounters (nearest neighbors). Statistically independent network segments are assumed to obey the Gaussian statistics. A recursive relationship for the distribution of chain lengths is developed, and it is shown that for large molecular. sizes this relationship tends to a limiting distribution. Corresponding average and standard deviation follow power law dependence on degree of polymerization (DP). With these results, cyclization probabilities are explicitly expressed as functions of DP. When segmental diffusion is the rate controlling factor, cyclization is similar to short range (or primary) cyclization. (c) 2005 Wiley Periodicals, Inc.