1. Characteristics of glycyl-L-leucine influx across the mucosal border of isolated guinea-pig ileum have been investigated. The influx of the peptide was measured with glycine-labelled or leucine-labelled compounds (Gly*-Leu or Gly-Leu*) and compared with that of a constituent amino acids under various experimental conditions 2. Gly-Leu* influx over a wide range of peptide concentrations. The latter obeyed simple Michaelis--Menten kinetics whereas the former could be described in terms of two saturable components. 3. Total replacement of medium Na with mannitol had no effect on Gly*-Leu influx, while it markedly reduced Gly-Leu* influx to a level slightly greater than Gly*-Leu influx. L-Leucine influx was partially dependent on Na in contrast to glycine influx which was absolutely dependent on Na. 4. Gly*-Leu influx was not inhibited by the simultaneous presence of glycine or L-isoleucine, while Gly-Leu* influx was strongly inhibited by L-leucine and L-isoleucine. Gly-Leu* influx under submaximal inhibition by L-isoleucine was about the same as Gly*-Leu influx. Di- or tri-glycine did not inhibit glycyl-L-leucine influx, while glycyl-L-leucine markedly inhibited diglycine influx, the inhibition being not competitive but of the mixed type. 5. A Michaelis--Menten type relation was observed for the increment in the transmural potential induced by glycyl-L-leucine, L-leucine or the mixture of the dipeptide and L-leucine. In all cases, the values of the maximum potential change were identical, suggesting that a single electrogenic transfer mechanism was operating in these cases. 6. It is concluded that about a half of glycyl-L-leucine influx is mediated by a carrier system for intact glycyl-L-leucine which is independent of sodium, and the other half is transported as L-leucine after membrane surface hydrolysis, part of this component being sodium-dependent and electrogenic. It is also suggested that the carrier sites for glycyl-L-leucine and glycylglycine are very closely located but separated.