The aim of this work is to study whether a quadrupole time-of-flight (QToF) mass analyzer, coupled to an ultra high performance liquid chromatography (UHPLC) system, can be a valuable alternative for a triple-quadrupole (QqQ) mass analyzer, for quantitative toxicological purposes. The case study considered was the quantification of 16 opioids (6-monoacetylmorphine, buprenorphine, codeine, dihydrocodeine, ethylmorphine, fentanyl, hydrocodone, hydromorphone, morphine, norbuprenorphine, norcodeine, norfentanyl, oxycodone, oxymorphone, pholcodine and tilidine) in human plasma. Both methods were validated in parallel in terms of selectivity, matrix effects, extraction recovery, carry-over, bias, precision and sensitivity. Accuracy-profile methodology was used to determine the optimal calibration model, and to estimate bias, repeatability, intermediate precision and total error. Selectivity was demonstrated for all opioids and deuterated analogues, except for codeine-d3 on the UHPLC-QTOF. For most compounds, extraction recoveries were in the range 60 to 80% on both systems, except for the synthetic analogues, buprenorphine, fentanyl and tilidine, where large variability is observed. Carry-over was negligible on both systems. For different opioids, the optimal calibration model was different between the systems. The accuracy profiles of the majority of the opioids indicated that, over the entire tested concentration range, for more than 5% of the future measurements, total errors are expected to exceed the a priori defined 15% acceptance limit. For some exceptions, however, the measurements even suffer from total errors above 30%, which can be attributed to the solid phase extraction procedure that was applied as sample pretreatment technique. Sensitivity was generally tenfold better on the LC-QToF system, probably due to the difference in ion choice for quantification between both systems. In conclusion, the best performing system seemed to depend on the compound, on the parameter and even on the concentration. Accuracy profiles clearly provided valuable information complementary to that obtained in classical validation tests, and therefore preferably are taken into account when deciding on a method's performance.