In this work, maximal α-leakage is introduced to quantify how much a quantum adversary can learn about any sensitive information of data upon observing its disturbed version via a quantum privacy mechanism. We show that the maximal expected α-gain of quantum measurement, α-leakage, and maximal α-leakage for a quantum privacy mechanism are characterized by measured conditional Rényi entropy, measured Arimoto information, and measured Rényi capacity, respectively. Various properties of maximal α-leakage, such as data processing inequality, composition property, and super-additivity for a memoryless quantum privacy mechanism, are established as well. Moreover, we discuss the asymptotic behavior of α-leakage and maximal α-leakage for quantum privacy mechanisms in the identically and independently distributed setting, and further prove their equivalence to sandwiched Rényi information and sandwiched Rényi capacity, respectively.