A new quantitative characterization of landscape-forming processes in the general framework of self-organized criticality and of fractal analyses is proposed. The coupled processes considered are threshold-independent hillslope evolutions and threshold-dependent fluvial transport phenomena. From a body of experimental and theoretical evidence we argue that geomorphological thresholds, principles of minimum energy expenditure and concepts of self-organized criticality are of crucial importance for the understanding of the basic general mechanisms which govern landscape evolution. This paper considerably extends both the theoretical framework and the empirical evidence for a recently developed theory which incorporates the above general principles. The modeling of landscape evolution by principles of self-organization is accomplished through the introduction of diffusion processes operating mainly on the hillslopes and the coupling of these processes with the fluvial evolution of the network previously studied through principles of self-organized criticality. The effects of spatial variability of surface erodibility are investigated under the general framework of random space functions with a correlation structure. Finally, a fractal analysis of the characteristics of the resulting landscape is performed and compared with recent results from real landforms to suggest the relationship of landscape fractal dimensions with the underlying landscape-forming processes.