Abstract. In this paper, we develop a sparse grid discontinuous Galerkin (DG) scheme for transport equations and applied it to kinetic simulations. The method uses the weak formulations of traditional Runge-Kutta DG (RKDG) schemes for hyperbolic problems and is proven to be L 2 stable and convergent. A major advantage of the scheme lies in its low computational and storage cost due to the employed sparse finite element approximation space. This attractive feature is explored in simulating Vlasov and Boltzmann transport equations. Good performance in accuracy and conservation is verified by numerical tests in up to four dimensions.Key words. discontinuous Galerkin methods; sparse grid; high-dimensional transport equations; Vlasov equation; Boltzmann equation.1. Introduction. In this paper, we develop a sparse grid DG method for high-dimensional transport equations. High-dimensional transport problems are ubiquitous in science and engineering, and most evidently in kinetic simulations where it is necessary to track the evolution of probability density functions of particles. Deterministic kinetic simulations are very demanding due to the large computational and storage cost. To make the schemes more attractive comparing with the alternative probabilistic methods, an appealing approach is to explore the sparse grid techniques [6,15] with the aim of breaking the curse of dimensionality [4]. In the context of wavelets or sparse grid methods for kinetic transport equations, we mention the work of using wavelet-MRA methods for Vlasov equations [5], the combination technique for linear gyrokinetics [19], sparse adaptive finite element method [25], sparse discrete ordinates method [16] and sparse tensor spherical harmonics [17] for radiative transfer, among many others. This paper focuses on the DG method [13], which is a class of finite element methods using discontinuous approximation space for the numerical solution and the test functions. The RKDG scheme [14] developed in a series of papers for hyperbolic equations became very popular due to its provable convergence, excellent conservation properties and accommodation for adaptivity and parallel implementations. Recent years have seen great growth in the interest of applying DG methods to kinetic systems (see for example [3,18,20,9, 10]) because of the conservation properties and long time performance of the resulting simulations. However, the DG method is still deemed too costly in a realistic setting, often requiring more degrees of freedom than other high order numerical calculations.Recently, we developed a sparse grid DG method for high-dimensional elliptic problems [24]. A sparse DG finite element space has been constructed, reducing the degrees of freedom from the standardwhere h is the uniform mesh size in each dimension. The resulting scheme retains main properties of standard DG methods while making the computational cost tangebile for high-dimensional simulations. This motivates the current work for the transport equations, and we use kinetic problems as a...