The tradeoffs between ballistic imaging (time-gated imaging of first-arrival, unscattered photons) and conventional imaging for resolving tumors in biological scattering media are examined. For ballistic imaging, closed form expressions are derived to characterize the resolvability using five degrees of freedom (laser intensity, scattering coefficient, thickness of medium, false alarm rate, and number of observations). For conventional imaging, a numerical approximation is used to find the asymptotic resolution using the scattering and absorption coefficients of the medium. Using the characterizations of both approaches, a decision-theoretic approach to determining the minimum resolvable object size is developed, which provides clear guidelines as to when time-gated ballistic imaging methods offer advantages over conventional imaging. The theoretical predictions are validated through a realistic simulation of tumors in breast tissue.