The reverse transformation to austenite is an important part in the heat treatment process of steels. The final microstructures obtained by quenching, normalization or intercritical annealing (e.g. for dual phase steels) are strongly dependent upon the reverse transformation kinetics and austenitic structure. Therefore, it is of considerable technical interest to study and control the kinetics of this transformation. In order to clarify effects of substitutional alloying elements on the kinetics of reverse transformation from pearlite, an Fe-0.6C binary alloy and Fe-0.6C-1 or 2M (M=Mn, Si) ternary alloys were used in the present study (hereafter denoted as 0.6C, 1Mn, 2Mn and 1Si alloys, respectively). The initial pearlite structure of each alloy has nearly the same interlamellar spacing of 0.17 μm. Thin sheets of 0.5 mm×5 mm×10 mm in size cut from the pearlitic specimen were firstly preheated in a salt bath at 923 K for 15 s, and then rapidly moved into another salt bath at 1073 K\linebreak for reverse transformation for various periods, followed by water quenching. The variation of Vickers hardness and microstructure evolution with holding time at 1073 K were examined. Ferrite region without cementite remains inside or neighboring to austenite newly formed during reverse transformation in the 0.6C and 1Si pearlitic specimens. It is indicated that continuous dissolution of cementite in ferrite and continuous diffusion of C atoms through ferrite into the growth front of austenite occur. However, pearlitic ferrite and cementite almost simultaneously disappeared in the 1Mn and 2Mn specimens. Reverse transformation from pearlite in the 0.6C pearlitic specimen at 1073 K is finished after holding for about 5.5 s. The reversion kinetics is retarded by the addition of Si. On the other hand, the reversion kinetics is slightly accelerated by the addition of Mn. The difference of acceleration between the 1Mn and 2Mn pearlitic specimens is small. From the viewpoint of 1D austenite growth in a pearlite lamella, austenite can grow without long range diffusion of Mn and Si in the Mn- and Si-added pearlitic specimens, respectively. It is supposed that austenite growth is controlled by carbon diffusion in those specimens. The addition of Mn or Si affects carbon diffusion by affecting carbon concentrations at the austenite/ferrite and austenite/cementite interfaces, resulting in changes in reversion kinetics.