We examine the prospects for supersymmetry discovery in the minimal supergravity (mSUGRA) model via indirect detection of neutralino dark matter. We investigate rates for muon detection in neutrino telescopes, and detection of photons, positrons and anti-protons by balloon and space based experiments. We compare the discovery reach in these channels with the reach for direct detection of dark matter, and also with the reach of collider experiments such as Fermilab Tevatron, CERN LHC and a $\sqrt{s}=0.5-1$ TeV linear $e^+e^-$ collider. We pay particular attention to regions of model parameter space in accord with recent WMAP results on the dark matter density of the universe. We find that the IceCube neutrino telescope as well as 3rd generation direct dark matter detection experiments should be able to cover the {\it entire} WMAP allowed portion of the hyperbolic branch/focus point (HB/FP) region of parameter space. This is in contrast to the case of the CERN LHC or a linear $e^+e^-$ collider, where only a fraction of the HB/FP region can be accessed. In addition, we show that detection of $\gamma$s, $e^+$s and $\bar{p}$s should occur in much of the HB/FP region, as well as in the low $m_{1/2}$ portion of the $A$ annihilation funnel, and will be complementary to searches via colliders in these regions.

We evaluate neutralino-nucleon scattering rates in several well-motivated supersymmetric models, and compare against constraints on the neutralino relic density, BF( b\to s\gamma ) as well as the muon anomalous magnetic moment a_\mu . In the mSUGRA model, the indirect constraints favor the hyperbolic branch/focus point (HB/FP) region of parameter space, and in fact this region is just where neutralino-nucleon scattering rates are high enough to be detected in direct dark matter search experiments! In Yukawa unified SUSY SO(10) models with scalar mass non-universality, the relic density of neutralinos is almost always above experimental bounds, while the corresponding direct detection rates are below experimental levels. Conversely, in five dimensional SO(10) models where gauge symmetry breaking is the result of compactification of the extra dimension, and supersymmetry breaking is communicated via gaugino mediation, the relic density is quite low, while direct detection rates can be substantial.