We show that, up to order type isomorphism, there is a unique crossing-minimal rectilinear drawing of K 18 . It is easily verified that this drawing does not contain any crossingminimal drawing of K 17 . Therefore this settles, in the negative, the following question from Aichholzer and Krasser: is it true that, for every integer n ≥ 4, there exists a crossingminimal drawing of K n that contains a crossing-minimal drawing of K n−1 ?