A series of model experiments which were designed to simulate certain aspects of the development of large-scale folds in bedded sedimentary rocks are described and illustrated. The models used were multilayers made up of thin sheets of gelatine. Two types of experiment were carried out. In the first a section of the base of a horizontally layered model was slowly raised to simulate uplift and gravity gliding. In the second a horizontally layered model was laterally compressed by a slow moving piston. The initial development of lateral compressive stress in both types of model was achieved by the propagation of a stress front along the length of a model. In a number of experiments this stress front played an important role in localizing the initiation of folds. The existence and rate of propagation of the stress front were related to the presence of shear stresses along the base of a model. The development of folds in the majority of experiments resulted in well defined anticlines separated by broad, poorly defined synclinal zones. Individual anticlines were initiated at different times during the course of an experiment and this resulted either in the serial development of a train of folds or in the development of isolated anticlines. Asymmetric anticlines were, during the initial stages of amplification, upright symmetrical structures. Asymmetry tended to develop late in the amplification history of such folds. In three dimensions, folds tend to be non-cylindrical and adjacent structures form an array of en echelon periclines.