We investigate whether future measurements of high-redshift standard candles will be a powerful probe of dark energy, when compared to other types of planned dark energy measurements. Active galactic nuclei, gamma-ray bursts, and certain types of core collapse supernova have been proposed as potential candidates of such a standard candle. Due to their high luminosity, they can be used to probe unexplored regions in the expansion history of the Universe. Information from these regions can help constrain the properties of dark energy, and in particular, whether it varies over time. We consider both linear and piecewise parameterizations of the dark energy equation of state, w(z), and assess the optimal redshift distribution that a highredshift standard-candle survey could take to constrain these models. The more general the form of the dark energy equation of state w(z) being tested, the more useful high-redshift standard candles become. For a linear parametrization of w(z), they give only small improvements over planned supernova and baryon acoustic oscillation measurements; a wide redshift range with many low-redshift points is optimal to constrain this linear model. However, to constrain a general, and thus potentially more informative, form of w(z), having many high-redshift standard candles can significantly improve limits on the nature of dark energy, even compared to dark energy experiments currently only in the planning stages.