We discuss a rate-theory model of polymer crystallization that exhibits many of the morphological and kinetic features observed in the laboratory. The interplay between the growth of the lamellar crystals in directions along the molecular direction and at right angles to it is discussed, together with some of the implications for the temperature dependence of the thickness of the crystals. We examine in some detail the hypothesis that the growth process selects the thickness that maximizes the rate of advance of the lamellar edge.