A non-zero cosmological constant is only one of many possible explanations for the observed accelerating expansion of the Universe. Any smoothly distributed, “dark” energy with a significant negative pressure can drive the acceleration. One possible culprit is a dynamical scalar field, but there are many less popular models such as tangled cosmic strings or domain walls. Soon theorists are likely to think up a number of new energies that can accelerate the expansion, meaning that only better observations can solve this question. Dark energy can be parameterized by its equation of state, w = p/ρ, which in the most general form can vary over time. Unlike the CMB, supernova observations cover a range of redshift so they can, in principle, probe the variation in the equation of state of the unknown component. The current SN observations loosely constrain the equation of state to w < −0.6, ruling out non-intercommuting strings and textures (w = −1/3), but consistent with a cosmological constant (w = −1). The constraints achievable from future large SN surveys are limited by our ability to understand systematic effects in SN Ia luminosities. But a large sample of supernovae reaching out to z ˜ 2 should at least discriminate between a cosmological constant and a dynamical scalar field as the source of the observed acceleration.