In Trinidad and Tobago, and the wider Anglophone Caribbean gender differences in academic performance is a major and complex topic. Unsurprisingly, Caribbean researchers have posited a number of explanations for gender variations in academic performance many of which are socio-cultural or economic in nature. However, research shows that the items on a scale may unintentionally produce bias towards either males or females and this may significantly affect the students? academic performance. Therefore, detecting potential sources of bias in scale items is important to ensure that scores derived from the scale are a valid and accurate representation of students? performance. In this paper, we use item response theory (IRT) to investigate the extent to which gender accounts for the patterns of responses of Trinidad and Tobago students on the Program for International Student Assessment (PISA) 2009. Specifically, we measure gender differential item functioning (DIF) in students? responses to the reading, mathematics, and science items on the assessment. Additionally, we propose three hypotheses for this study: 1) Reading items will show DIF in favor of females, 2) Mathematics items will show DIF in favor of males, and 3) Science items will show DIF in favor males. This study contributes to research on assessment and evaluation by demonstrating how the content of a scale and students? responses to items on a scale can provide valuable information about student performance. This presentation is suited to the ?Transforming Assessment and Evaluation? category.