I present examples of plant functional-structural models (FSMs) that are used to evaluate how foliage units affect whole-canopy functions, and I show that multi-criteria optimization is an effective tool for these models. FSMs produce plant structures through the repeated application of a set of rules for the placement of foliage units. The models are blind (rules are the same regardless of dynamic simulation conditions), sighted (rules change with interference from other foliage units) or self-regulatory (rules change depending on the conditions of the simulation, i.e., internal conditions). In the examples presented, the models are used to optimize plant morphology for one or more measures of plant performance; these measures include movement of materials and associated hydraulic functions, foliage display, light interception and net carbon, mechanical support and stability, and reproductive success. It is consistently found that no morphology is optimal for any single measure of plant performance, and the rules for plant development are not stationary in space and time. In multi-criteria optimization, alternative morphologies are compared against multiple measures of plant performance; these are optimized simultaneously using Pareto optimality, which yields the set of mutually co-dominant solutions not dominated by any other solution. Two solutions are considered to be mutually co-dominant if improvement with respect to one criterion is at the expense of another criterion. I conclude that multi-criteria optimization is an essential tool for the use of FSMs to relate processes at the foliage level to whole-canopy function and to explain the structural diversity of old-growth forests.