This paper proposes a method for obtaining new information on 3D radiative effects that arise from horizontal radiative interactions in heterogeneous clouds. Unlike current radiative transfer models, it can not only calculate how 3D effects change radiative quantities at any given point, but can also determine which areas contribute to these 3D effects, to what degree, and through what mechanisms. The new method uses Monte Carlo radiative transfer simulations to generate numerous photon trajectories through the cloud field, and then it examines the radiative processes along each trajectory. After describing the proposed method, the paper illustrates its new capabilities both for detailed case studies and for the statistical processing of large datasets. Because the proposed method makes it possible, for the first time, to link a particular change in cloud properties to the resulting 3D effect, it can be used to develop new types of radiative transfer parameterizations. Encouraging initial results suggest that such parameterizations will be able to incorporate 3D effects in practical applications currently limited to 1D theory--such as remote sensing of cloud properties and dynamical cloud modeling.