In 1983, in the Netherlands' highly differentiated school system, two types of curriculum representing different ability levels were combined as a first step towards a more heterogeneous grouping of student abilities. A study of one aspect of the results of this change compared over 1000 samples of English and German second language listening comprehension tests administered before and after the change. Statistical analysis of test results from students representing two different ability levels in different years indicate that the curriculum change may have had a negative effect at both levels of achievement. The results indicate that mixing ability levels has a negative effect on the final ability attained. The findings are supported by results on tests of other foreign language skills (reading comprehension) and in another subject (chemistry). (MSE)