Generational intelligence gains are one intriguing finding in science. Nutrition and cognitive stimulation are among the most remarkable causes of the upward trend in intelligence. The nutrition hypothesis predicts a primary impact on the most deprived, producing disproportionate gains at low intelligence levels. The cognitive stimulation hypothesis predicts gains along the intelligence distribution. However, data from the entire distribution are rarely available. The present study compares a sample of children tested in 1970 with an equivalent sample tested 30 years later. Data for the entire distributions were available. The results are consistent with the nutrition hypothesis, because the gains were mainly concentrated in the lower and medium halves of the distribution and were negligible in the very top half of the distribution. Moreover, an impressive gradual decrease in the gains was observed from the lower half to the top half of the distribution.