Measurements of the macroscopic growth rates of the (101) face of tetragonal lysozyme crystals indicate an unusual dependence on the supersaturation [Forsythe et al. (1999), Acta Cryst. D55, 1005-1011] similar to that observed for the (110) face. As performed previously for the (110) face, the surface packing arrangement for the (101) face was constructed in this study based on earlier microscopic observations and theoretical analysis of the internal molecular packing. This allowed the minimum growth unit for this face to be identified as a tetramer corresponding to a single turn of helices centered about the 43 axes and the minimum growth step to be identified as of unimolecular height. A macroscopic mathematical model for the growth of the (101) face was developed based on the reversible formation of multimeric growth units in solution and the addition of a unit to the crystal face by dislocation and two-dimensional nucleation mechanisms. The calculations showed that the best fits were obtained for tetramer or octamer growth units in this model. This and other evidence suggests that while growth may proceed by a variety of growth units, the average size of these units is between that of a tetramer and an octamer.