A set of hydrophilic fluorescent dyes of known molecular weight has been used to determine the molecular exclusion limit and the extent of apical, epidermal and cortical symplasts in the root, stem and leaf of Egeria densa. These dyes are unable to pass the plasmalemma, so that any cell-to-cell movement of injected dye must occur via the symplast. The shoot-apex symplast has a high molecular exclusion limit, excluding dyes with a molecular weight of 749 dalton (fluorescein hexaglycine) and greater but allowing dyes of up to 665 dalton (fluorescein diglutamic acid) to pass. The leaf epidermal symplast is similar to that in the apex: fluorescein pentaglycine (674 dalton) moves to a limited extent, but fluorescein hexaglycine is immobile. Stem and root epidermal cells have a lower molecular exclusion limit, only the dye 6-carboxyfluorescein (376 dalton) is able to move from cell-to-cell. Cortical and epidermal tissues in both the stem and the root have similar symplast permeabilities. However, a barrier to dye (6-carboxyfluorescein) movement is found between the epidermis and the cortex in both organs. Barriers are also found at the nodes between expanded internodes. The stem barriers are not found in the unexpanded nodes near the shoot tip; apparently they are formed early during internode expansion. In the root tip, a barrier to the movement of dye is found between the root cap and the remainder of the root. Plasmodesmata are found linking all cell types studied, even cells where barriers to dye movement occur. Thus, the plant, far from being one uniform symplast, consists of a large number of symplast domains, which may or may not differ in molecular exclusion limit.