The need for long-term predictions is pointed out, direction into long-term time scales is defined. Spatial organization, solar activity and their implications for predictions are discussed. A definition of solar minimum is given and five methods of solar cycle prediction are briefly outlined. Some recommendations are given.