In this study, we examined the presence of differential item functioning (DIF) among groups of students who were tested in their native language or in a different language when participating in the 1999 Trends in International Mathematics and Science Study. Data from 18,837 examinees from three countries (Singapore, United States, and Iran) were analyzed. Logistic regression DIF detection procedures were used in conjunction with effect size criteria, and repeated random samples were drawn to evaluate the effect of sample size differences across groups. Only one item was flagged for non-negligible DIF, suggesting that for the countries studied there was little substantive DIF due to testing in a non-dominant language. Differences across reference and focal group sample sizes did not affect overall conclusions of DIF. Suggestions for future research in this area are provided. (Contains 4 figures, 4 tables and 2 notes.)