In this paper we compute the the one-loop chiral logarithmic corrections to
the S and T parameters in a highly deconstructed Higgsless model with only
three sites. In addition to the electroweak gauge bosons, this model contains a
single extra triplet of vector states (which we denote \rho^{\pm} and \rho^0),
rather than an infinite tower of "KK" modes. We compute the corrections to S
and T in 'tHooft-Feynman gauge, including the ghost, unphysical
Goldstone-boson, and appropriate "pinch" contributions required to obtain
gauge-invariant results for the one-loop self-energy functions. We demonstrate
that the chiral-logarithmic corrections naturally separate into two parts, a
model-independent part arising from scaling below the \rho mass, which has the
same form as the large Higgs-mass dependence of the S or T parameter in the
standard model, and a second model-dependent contribution arising from scaling
between the \rho mass and the cutoff of the model. The form of the universal
part of the one-loop result allows us to correctly interpret the
phenomenologically derived limits on the S and T parameters (which depend on a
"reference" Higgs-boson mass) in this three-site Higgsless model. Higgsless
models may be viewed as dual to models of dynamical symmetry breaking akin to
"walking technicolor", and in these terms our calculation is the first to
compute the subleading 1/N corrections to the S and T parameters. We also
discuss the reduction of the model to the ``two-site'' model, which is the
usual electroweak chiral lagrangian, noting the ``non-decoupling''
contributions present in the limit as M_\rho goes to infinity.