The kinetics of isoleucine, leucine, and valine transport in Escherichia coli K-12 has been analyzed as a function of substrate concentration. Such analysis permits an operational definition of several transport systems having different affinities for their substrates. The identification of these transport systems was made possible by experiments on specific mutants whose isolation and characterization is described elsewhere. The transport process with highest affinity was called the "very-high-affinity"process. Isoleucine, leucine, and valine are substrates of this transport process and their apparent K(m) values are either 10(-8), 2 x 10(-8), or 10(-7) M, respectively. Methionine, threonine, and alanine inhibit this transport process, probably because they are also substrates. The very-high-affinity transport process is absent when bacteria are grown in the presence of methionine, and this is due to a specific repression. Methionine and alanine were also found to affect the pool size of isoleucine and valine. Another transport process is the "high-affinity" process. Isoleucine, leucine, and valine are substrates of this transport process, and their apparent K(m) value is 2 x 10(-6) M for all three. Methionine and alanine cause very little or no inhibition, whereas threonine appears to be a weak inhibitor. Several structural analogues of the branched-chain amino acids inhibit the very-high-affinity or the high-affinity transport process in a specific way, and this confirms their existence as two separate entities. Three different "low-affinity" transport processes, each specific for either isoleucine or leucine or valine, show apparent K(m) values of 0.5 x 10(-4) M. These transport processes show a very high substrate specificity since no inhibitor was found among other amino acids or among many branched-chain amino acid precursors or analogues tried. The evolutionary significance of the observed redundancy of transport systems is discussed.