A zonally averaged, physically based sea ice model is coupled to an energy balance climate model with prescribed observed atmospheric temperatures and used to investigate the sea ice model sensitivity to alternative parameterizations of (1) the upward heat flux Fb to the ice base, and (2) lateral melting of ice from open water. The parameterizations tested here are taken from the climate and sea ice modeling literature. The different parameterizations for each of these processes can lead to dramatically different results in simulated sea ice area, sea ice thickness, or both. Parameterizing Fb by prescribing (or computing) an upward heat flux at the mixed layer base and either constraining the subice mixed layer temperature to equal the freezing point of seawater or parameterizing Fb as a function of subice mixed layer temperature has the advantage of incorporating the effects on heat fluxes to the ice base of both convective overturning and meridional oceanic heat flux convergence. Parameterizing lateral melting as a function of lead temperature Tlead, rather than making it depend on ∂Tlead/∂t, has relatively little impact on simulated winter ice extent, and no impact on summer ice extent. ¿ American Geophysical Union 1990