Oral fluid (OF) offers a noninvasive sample collection for drug testing. However, 3,4-methylenedioxymethamphetamine (MDMA, ecstasy) in OF has not been adequately characterized in comparison to plasma. We administered oral low-dose (1.0 mg/kg) and high-dose (1.6 mg/kg) MDMA to 26 participants and collected simultaneous OF and plasma specimens for up to 143 h after dosing. We compared OF/plasma (OF/P) ratios, time of initial detection (t first), maximal concentrations (C max), time of peak concentrations (t max), time of last detection (t last), clearance, and 3,4-methylenedioxyamphetamine (MDA)-to-MDMA ratios over time. For OF MDMA and MDA, C max was higher, t last was later, and clearance was slower compared to plasma. For OF MDA only, t first was later compared to plasma. Median (range) OF/P ratios were 5.6 (0.1-52.3) for MDMA and 3.7 (0.7-24.3) for MDA. OF and plasma concentrations were weakly but significantly correlated (MDMA: R(2) = 0.438, MDA: R(2) = 0.197, p < 0.0001). Median OF/P ratios were significantly higher following high dose administration: MDMA low = 5.2 (0.1-40.4), high = 6.0 (0.4-52.3, p < 0.05); MDA low = 3.3 (0.7-17.1), high = 4.1 (0.9-24.3, p < 0.001). There was a large inter-subject variation in OF/P ratios. The MDA/MDMA ratios in plasma were higher than those in OF (p < 0.001), and the MDA/MDMA ratios significantly increased over time in OF and plasma. The MDMA and MDA concentrations were higher in OF than in plasma. OF and plasma concentrations were correlated, but large inter-subject variability precludes the estimation of plasma concentrations from OF.