Background: Shayer, Ginsburg, and Coe (2007) showed that children leaving primary school in Y6 entered secondary school with much lower levels of understanding of the physical conservations than in 1976. It seemed desirable to investigate cognitive development in the first three years of secondary education. Aims: By using two Piagetian tests of formal operations, one of which had been used in the 1976 CSMS survey, the performance of Y8 and Y9 pupils would be compared with the survey results published in 1978. Sample: Eight schools were selected as willing to test their Y9 or Y8 classes on either the "Pendulum" (in 2007) or the "Equilibrium in the Balance" tests (in 2006), giving 39 classes on which to make the comparison with 1976 performance. Method: Regression of the pupils' class mean on each formal test on either the class's 2004 MidYIS or nferCAT standardized scores, and computing the regression at IQ = 100 allows comparison with that found in 1976. Results: On "Equilibrium in the Balance" the Y8 pupils were down on the proportion at the early formal level or above by -0.61 standard deviations (SD) for boys and -0.63 SD for girls on the 1976 results. On the "Pendulum" the Y9 boys were down by -0.49 SD and the girls by -0.48 SD on the proportion at the early formal level or above. Conclusion: The negative Flynn-effect found on "Volume & Heaviness" for Y7 pupils is paralleled by a similar negative effect on attainment of formal operations by Y8 and Y9, compared with 1976. Yet at the same time the proportion of pupils using the top level of concrete operational thinking has increased on both tests. It seems that there has been a change either in general societal pressures on the individual or in the style of teaching in schools--or both--favouring a lower level of processing of reality.