One-dimensional snowpack algorithms in major operational snowmelt models used in the United States (HEC-1, SSARR, NWSRFS, SRM, and PRMS) are reviewed and contrasted with two U.S. Army Cold Regions Research and Engineering Laboratory (CRREL) algorithms (SNTHERM and SNAP) that are candidates for use in distributed operational models. In contrast to current operational models, the CRREL algorithms provide more detail in snowpack processes and require no calibration. The CRREL algorithms also include a full surface energy balance that requires more meteorological data than most operational models. Simpler surface energy balances could be used with the CRREL models. In future modeling systems, it would be preferable for the surface energy balance algorithms to be made independent of the internal snowpack process algorithms, so that available meteorological data can be used to drive a snowpack model of choice. Improvements are needed in the way that forest canopies and other groundcovers are accounted for in the surface energy balances of the CRREL models.