The collection and analysis of drugs in oral fluid (OF) at the roadside has become more feasible with the introduction of portable testing devices such as the Alere™ DDS®2 Mobile Test System (DDS®2). The objective of this study was to compare the on-site results for the DDS®2 to laboratory-based confirmatory assays with respect to detection of drugs of abuse in human subjects. As part of a larger Institutional Review Board approved study, two OF samples were collected from each participant at a music festival in Miami, FL, USA. One OF sample was field screened using the DDS®2, and a confirmatory OF sample was collected using the Quantisal™ OF collection device and submitted to the laboratory for testing. In total, 124 subjects participated in this study providing two contemporaneous OF samples. DDS®2 field screening yielded positive results for delta-9-tetrahydrocannabinol (THC) (n = 27), cocaine (n = 12), amphetamine (n = 3), methamphetamine (n = 3) and benzodiazepine (n = 1). No opiate-positive OF samples were detected. For cocaine, amphetamine, methamphetamine and benzodiazepines, the DDS®2 displayed sensitivity, specificity and accuracy of 100%. For THC, the DDS®2 displayed sensitivity of 90%, specificity of 100% and accuracy of 97.5%, when the threshold for confirmation matched that of the manufacturers advertised cut-off. When this confirmatory threshold was lowered to the analytical limit of detection (i.e., 1 ng/mL), apparent device performance for THC was poorer due to additional samples testing positive by confirmatory assay that had tested negative on the DDS®2, demonstrating a need for correlation between manufacturer cut-off and analytical reporting limit. These results from drug-using subjects demonstrate the value of field-based OF testing, and illustrate the significance of selecting an appropriate confirmation cut-off concentration with respect to performance evaluation and detection of drug use.