Academic dishonesty has become a pressing challenge in higher education, particularly with the expansion of online and blended learning. Using a cross-sectional survey, this study examined the prevalence and determinants of academic dishonesty among undergraduate students in a Malaysian higher learning institution, focusing on demographic predictors such as gender, year of study, and academic performance, in the digital age. Descriptive and inferential analyses, including cross-tabulations, t-tests, and multiple regression, were used to explore patterns of dishonest behaviors. The results indicate that male students were more likely to engage in high-risk cheating behaviors, such as copying answers during exams, while female students showed slightly higher involvement in collaborative dishonesty, such as sharing answers with peers. First and second-year students reported higher levels of dishonesty compared to seniors, suggesting that academic maturity reduces engagement in misconduct. Students with lower CGPAs (2.00–2.99) displayed significantly higher dishonest behaviors than those with CGPAs above 3.50. Regression analysis confirmed that gender, year of study, and academic performance jointly predicted dishonesty, though the model explained some variance, indicating other influential factors. The study highlights the importance of targeted interventions for early-year and academically struggling students, as well as the need for innovative assessment strategies that safeguard integrity in blended and online environments.