Introduction: Tuberculosis (TB) remains the leading cause of death among people living with HIV (PLHIV). In intermediate-burden settings, TB prevention may be improved by using simple predictors routinely available in HIV care. We evaluated CD4 cell count and plasma HIV viral load (VL) as predictors of active TB among PLHIV in Morocco. Methods: we conducted a retrospective cohort study at the Centre of Virology, Infectious and Tropical Diseases (CVMIT), Mohammed V Military Teaching Hospital, Rabat, Morocco. Patients registered between March 2016 and December 2023 with at least 2 years of follow-up and complete immunovirological data were included. Active TB was defined as initiation of antituberculous treatment during follow-up, with microbiological confirmation and/or compatible radiological findings, when documented. Receiver operating characteristic (ROC) curve analysis was used to assess discrimination, and optimal cutoff values were determined using Youden´s index. Results: a total of 132 participants were included (72.7% male; mean age, 47.1 ± 12.2 years), of whom 33 (25.0%) developed active TB. CD4 count showed modest discriminatory performance (AUC, 0.609), with an optimal threshold of 316 cells/µL (sensitivity, 43.0%; specificity, 76.7%; positive predictive value, 30.3%; negative predictive value, 85.1%). VL performed better (AUC, 0.708), with an optimal cutoff of 692 copies/mL (sensitivity, 77.4%; specificity, 58.4%; positive predictive value, 39.3%; negative predictive value, 88.1%). Among patients at the AIDS stage, CD4 count was not discriminatory (AUC, 0.508; p = 0.557), whereas VL remained significantly predictive (AUC, 0.662; p < 0.001). Conclusion: HIV viral load showed better predictive performance than CD4 count for TB risk stratification, particularly in patients with advanced HIV disease. Viral load-based stratification may help prioritise intensified TB screening and targeted preventive strategies in Moroccan PLHIV, alongside clinical evaluation and microbiological testing.