Studies in different countries, which have shown that rates of mobility are more alike than they were expected to be, have raised questions about the mechanisms underlying social mobility. The following paper puts forward the suggestion that social mobility belongs to a cybernetic mechanism whereby something like a ‘steady state’ is maintained in each occupational class by means of constant movement into and out of it. The quality discussed is ‘intelligence’; and the argument stems from two main observations. The first is that the higher the occupational class, the higher is the measured intelligence. The second is that the children of parents towards the extreme show a ‘regression’ from that extreme towards the mean of the general population. If these two observations hold true over a period of time, it follows that there must be social mobility. In a third part of the paper it is argued that differential fertility between the classes gives rise to additional mobility. The last section discusses longer‐term changes which may influence the amplitude of short‐term oscillations between any two generations, and finally notes a certain parallelism with the laws of