The effects of nonlinearity in several clamped clamped beam vibration problems are reviewed. Possible effects are: jump or instability phenomena, amplitude-limiting effects, and distortion of probability densities. Studies of prototype panel frame structures show similar behavior. The necessity for extending these efforts to built-up structures is emphasized. Also, estimates for the onset of nonlinearity based on two simple models are made to show how one can be guided in experimental design by quantitative-empirical considerations. Some theoretical techniques are appraised for their applicability to the structures problem. In particular, two approximation methods are singled out for detailed comment. Finally, the possibilities of a more creative use of experimental analysis and a closer tie between theoretical and experimental effort in structures research are explored.