This paper examines the most influential model of tourism development, the life cycle model, which has defined a ‘normal science’ of evolution for tourism destinations. The model is internally coherent, logical, easily intelligible and was used extensively as a guide to predicting development. For all destinations it postulates a path of steady growth until stagnation is reached. The model's widespread influence has discouraged policy and planning prior to stagnation. This paper illustrates the model's limitations, through a comparative analysis of tourism development in Queensland and Hawaii. Tourism growth in actual destinations does not support the model's predictions, in that some of the most attractive, but undeveloped destinations, have shown limited growth, while there are no indications of decline in established ‘mass’ destinations. Tourism growth in undeveloped areas is not inevitable, nor need developed destinations eventually stagnate. The paper argues that there is no automatic developmental progression. At all levels of development there are crisis points requiring coherent policy implementation if sustainable tourism growth is to be achieved.