A special biased Monte Carlo algorithm is used to study flow of homopolymer melts between neutral, hard walls on a fcc lattice at full occupancy (phi = 1). A random number biasing technique is developed to mimic slot flow of a melt; the biasing method preferentially moves the chains in the direction of flow. System properties including velocity profiles, chain-end density distributions, average radii of gyration, and end-to-end vector order parameters are investigated as functions of chain length and biasing parameter. Chain connectivity leads to non-Newtonian flow behavior evidenced as velocity profile blunting. Observation suggests a relationship between the logarithm of the biasing parameter and the pressure drop. Based on the prescribed biasing profile, it is shown that flow causes greater chain deformation for longer chains (N = 64,256) than for smaller chains (N = 16). Findings also include the ordering of chains with end-to-end vectors aligned with respect to the flow direction. (C) 2000 American Institute of Physics. [S0021-9606(00)50812-5].