This report discusses the relationship between bilingualism and mental development of bilingual children. After a review of the relevant literature, a specific study is described. The linguistic background of 648 children from 29 schools, age 10 through 12 inclusive, was measured with the Welsh Linguistic Background Scale. General intelligence was assessed with three nonverbal tests: Raven's Progressive Matrices, Daniel's Figure Reasoning Test and the Non-Verbal Test No. 2 of the National Foundation for Educational Research. Objectives were to: (1) determine which of the nonverbal tests was the most independent of linguistic background and hence most suitable for intelligence testing in mixed language areas, and (2) describe the relationship between linguistic background and test scores. Conclusions are that: (1) none of the tests used is completely suitable for evaluating Welsh-speaking children if the results are to be compared with those of English-speaking children; (2) Welsh-dominant bilingual children scored consistently lower than English-dominant children; (3) a correlation exists between test performance and degree of bilingualism; (4) the location of a community accentuates the influence of bilingualism; (5) occupational levels and socioeconomic status must be considered when interpreting test scores; and (6) Raven's Progressive Matrices is the most independent test. (CLK)