Conceptually, there should be a brief interval in time when newly secreted proteins "pile up" at secretory sites just outside the membrane of ameloblasts. Indeed, previous cytochemical studies have suggested that glycosylated and/or sulfated glycoproteins accumulate at enamel growth sites. Colloidal gold lectin cytochemistry and immunocytochemistry with antibodies to enamel proteins and phosphoserine, combined with cycloheximide and brefeldin A to inhibit protein synthesis and secretion, were applied to characterize the distribution of newly formed proteins at enamel interrod and rod growth sites. Although enamel growth sites show a "rarefied" appearance, the results indicate that one or more subclasses of enamel proteins accumulate near the cell surface at sites where elongation of enamel crystallites contributes to thickening of the enamel layer. These proteins are glycosylated and/or phosphorylated and, at least in the case of the glycosylated ones, are rapidly processed after they are released extracellularly. In contrast, immunolabeling for amelogenins is generally weaker near the cell surface and more intense at a short distance away from the site where crystallites elongate. The data suggest that the enamel proteins accumulating at growth sites likely belong to the non-amelogenin category and play a transient role in promoting the lengthening of crystallites. It is concluded that areas near the ameloblast membrane where certain enamel proteins accumulate in fact constitute the equivalent of a mineralization front.