SUMMARYOnly when the data base is of sufficient quality can a theoretical model of any process be justified. Such a data base has recently been assembled in the field of damping in tall buildings. It is now possible to specify a damping value appropriate to a particular building vibrating at a specified amplitude which is well correlated with values which are measured in practice. The mechanism on which this predictor is based suggests that the predictor will be applicable to other types of structure as well. As yet the data base for these other structures does not exist, but some circumstantial evidence is encouraging.