Drug-induced liver injury (DILI) can occur after overexposure to many different chemical compounds, including ingredients found in herbal remedies, nutrient supplements, and other over-the counter and prescription medicines. DILI is the most common cause of acute liver failures in the United States. In addition, DILI is a major cause for FDA disapproval of therapeutic drugs, black-box labeling, and complete withdrawal from the market. Generally it costs several hundred million dollars to bring a drug to the public. However, just one incidence of fatal liver injury can take the drug off the market and from the patients who really need it. Adequate animal models and effective screening assays are lacking for assessing DILI, especially when each drug may have different mechanism of action causing DILI. Based on the current research, it is thought that many DILI cases can arise from immune mediation by sensitized T cells when parent drug or its metabolites form hapten to serum protein or other membrane proteins and bind to the T-cell receptor to trigger immune response which leads to allergy as in hapten hypothesis. Also drugs can directly bind to the receptors and induce immune reaction as in P-I hypothesis. These immune-mediated DILI can be idiosyncratic, hard to predict and fatal. Hence, it is important to develop a diagnostic in vitro assay to predict immune-mediated DILI. The objective of this study was to develop an in vitro modified lymphocyte transformation test (MLTT) with IL-2, IL-5, IL-13, INF-_, and Granzyme B, using drug-allergic donors and healthy normal donors. Optimal PBMC isolation method, frozen PBMC thawing temperature, and time point was assessed. Lymphoprep method was the best in isolating PBMC and 37 °C quick- thaws produced higher yield of viable cells. Moreover, 4- day incubation was the most optimal for generating IL-2, IL-5, IL-13, INF-_, and Granzyme B data. Furthermore, intra- and inter-assay variability were addressed and found to be very high for MLTT. Finally, sensitivity and specificity of MLTT was determined to be 78 % and 44 % respectively by screening 14 drug-allergic donors and 9 healthy donors. Based on the high intra- and inter-assay variability with relatively low sensitivity and specificity, the MLTT measuring cytokines and Granzyme B did not meet the standard validation criteria. Nevertheless, this study demonstrated the usefulness of the MLTT for further exploring mechanisms of various drugs and metabolites involved in immune-mediated DILI.