We use Monte Carlo techniques to simulate the ability of future large high- redshift galaxy surveys to measure the temporal evolution of the dark energy equation of state w( z), using the baryonic acoustic oscillations in the clustering power spectrum as a `` standard ruler.'' Our analysis utilizes only the oscillatory component of the power spectrum and not its overall shape, which is potentially susceptible to broadband tilts induced by a host of model- dependent systematic effects. Our results are therefore robust and conservative. We show that baryon oscillation constraints can be thought of, to high accuracy, as a direct probe of the distance- redshift and expansion rate - redshift relations where distances are measured in units of the sound horizon. Distance precisions of 1% are obtainable for a fiducial redshift survey covering 10,000 deg(2) and redshift range 0: 5 < z < 3: 5. If the dark energy is further characterized by w(z) w(0) + w(1)z ( with a cutoff in the evolving term at z = 2), we can then measure the parameters w(0) and w(1) with a precision exceeding current knowledge by a factor of 10: 1 sigma accuracies Delta(w0) approximate to 0: 03 and Delta w(1) approximate to 0: 06 are obtainable ( assuming a flat universe and that the other cosmological parameters Omega(m) and h could be measured independently to a precision of +/- 0.01 by combinations of future CMB and other experiments). We quantify how this performance degrades with redshift/ areal coverage and knowledge of Omega(m) and h and discuss realistic observational prospects for such large- scale spectroscopic redshift surveys, with a variety of diverse techniques. We also quantify how large photometric redshift imaging surveys could be utilized to produce measurements of ( w(0); w(1)) with the baryonic oscillation method, which may be competitive in the short term.