We have developed a Monte-Carlo computer simulation to study associating polymer interactions. In our model we treat the associations as geometrical constraints. Each polymer chain contains two‘stickers’. The chains are treated as lattice selfavoiding random walks. Each sticker is constrained to be adjacent to one other sticker, but the stickers are free to exchange partners. This freedom to exchange results in an attraction between the chains, as anticipated by Cates and Witten.1 We find that in equilibrium the mutual excluded volume of two such chains passes from repulsive to attractive when the ratio of the sticker distance to the chain length is approximately 0.8. These results are independent of the chain length: they should apply to real polymers subject to these topological constraints in any good solvent at sufficiently high molecular weight.