The sizes and structures of isolated functionalised polymers in a hydrocarbon solvent are studied using atomistic molecular dynamics simulations and Monte Carlo simulations of coarse-grained chains. A speciﬁc functionalised polyethylene-polypropylene random copolymer in n-heptane is studied using atomistic simulations. The functional groups contain aromatic and polar groups, and 8 of them are distributed on an 8 kDa polymer backbone in several diﬀerent ways. It is shown that the radius of gyration and the end-to-end distance depend sensitively on the functionalgroup distribution. A random distribution of functional groups gives the most compact polymer structure, but other distributions gives values up to 50% larger; the largest values are when the functional groups are split evenly between both ends of the polymer. This is shown to be due to the association of the polar, and hence solvophobic, functional groups. A coarse-grained bead-spring model is then studied that includes solvophilic beads (representing unfunctionalised units) and solvophobic beads (representing functionalised units). Monte Carlo simulations are used to survey functional-group concentration and distribution. The results show that the collapse of a polymer with increasing solvophobicity depends sensitively on the distribution of diﬀerent beads. Form factors are presented for both the atomistic and coarse-grained models, and are analysed as if they were experimental scattering measurements. The apparent radii of gyration are in good agreement with those determined directly from simulation.