Previous studies based on population cohorts born at least 35 years ago, have reported appreciable childhood cognitive deficits for twins. We compared longitudinal IQ scores from approximately 188,000 singletons and some 6000 twins who went to primary school in the Netherlands from 1994 to 2003. In addition, we used a family-based design in which IQ scores of "adult" twins (N = 196) were compared with those of their adult singleton siblings (N = 589). After correcting for such confounding factors as the year of testing, gender, age at the time of the test, and parents' education and ethnicity, twins aged 6 scored 16% of a standard deviation lower than non-twins in language and 17% of a standard deviation lower in arithmetic. For twins aged 8 the difference with non-twins in language and arithmetic reduced to 5% and 2% of a standard deviation and for twins aged 10 and 12 the differences were not statistically significant. For IQ scores, twins scored 0.09 points lower than non-twins at age 8 and 0.83 points lower at 10. However, twins scored higher at age 12 by 0.14 points. The only significant difference found was at age 10. Using the family-based adult sample, no differences in IQ scores were found between twins and their singleton siblings. These results suggest that in a very recent generation of school children in the Netherlands, there was a small but significant cognitive deficit for twins aged 6 and 8. However, the difference disappeared by the time the children were 12, and was also insignificant in the adult population. Previous studies, based on cohorts born more than 35 years ago in Britain, reported much larger cognitive deficits in twins. Whatever the reason of the cognitive deficit at age 6, which could include prenatal growth, shorter gestation and parental care, twins caught up and the cognitive cost of being a twin in the Netherlands seems to be minor and temporary. (Contains 3 tables.)