In this study, we take a critical look at the degree to which the measurement of bullying and sense of belonging at school is invariant across groups of students defined by immigrant status. Our study focuses on the invariance of these constructs as measured on a recent PISA administration and includes a discussion of two statistical methods for assessing measurement invariance--multiple-group confirmatory factor analysis (MGCFA) and alignment optimization--. We discuss and illustrate how the alignment optimization method is optimal for handling data from large-scale international assessments like PISA. An acceptable degree of noninvariance was achieved for the two scales.