This study examines the mechanisms of epithelial cell rearrangement during archenteron elongation in the sea urchin embryo using scanning electron microscopy, differential interference contrast videomicroscopy, cell marking, and fluorescently labeled chimaeric clones. Archenteron elongation involves two major processes: local shifts in position of cells in the archenteron wall and polarized motility of the cells as they rearrange. Fluorescently labeled chimaeric clones introduced into the archenteron of Lytechinus pictus are initially 4-5 cells wide; by the end of gastrulation the clones elongate and narrow, so that they are one cell wide in the narrowest region of the archenteron. The extent of clonal mixing indicates that cells in the archenteron change their relative positions by only 1-2 cell diameters during cell rearrangement. Cells at the blastopore rearrange concomitantly with cells in the archenteron, resulting in a 35% decrease in blastopore diameter. Endoderm cells undergo polarized, stage-specific changes in shape and motility as they rearrange; (1) they flatten markedly along their apical-basal axis throughout archenteron elongation; (2) just prior to the onset of cell rearrangement, basal surfaces of all cells in the archenteron extend long, polarized lamellipodial protrusions along the axis of extension of the archenteron; (3) as cell rearrangement begins, basal surfaces round up and the cells become isodiametric; (4) by the 3/4 gastrula stage the cells become stretched along the animal-vegetal axis, apparently due to filopodial traction, and finally (5) they continue to rearrange, returning to a less elongated shape by the end of gastrulation. Direct observation of gastrulation in the cidaroid Eucidaris tribuloides indicates that in this species cell rearrangement is accomplished by progressive circumferential intercalation of cells without upwardly directed filopodia. This intercalation is accompanied by explosive, apparently stochastic, cortical blebbing activity at the boundaries between cells, suggesting that in addition to whatever cell rearrangement may be generated by filopodial tension, such activity is an important component of the active rearrangement process.