QTLs for salt-tolerance (ST) related traits at the seedling and tillering stages were identified using 99 BC(2)F(8) introgression lines (IL) derived from a cross between IR64 (indica) as a recurrent parent and Binam (japonica) from Iran as the donor parent. Thirteen QTLs affecting survival days of seedlings (SDS), score of salt toxicity of leaves (SST), shoot K(+) concentration (SKC) and shoot Na(+) concentration (SNC) at the seedling stage and 22 QTLs underlying fresh weight of shoots (FW), tiller number per plant (TN) and plant height (PH) at the tillering stage were identified. Most QTLs detected at the tillering stage showed obvious differential expression to salt stress and were classified into three types based on their differential behaviors. Type I included 11 QTLs which were expressed only under the non-stress condition. Type II included five QTLs expressed in the control and the salt stress conditions, and three of them (QPh5, QPh8 and QTn9) had similar quantity and the same direction of gene effect, suggesting their expression was less influenced by salt stress. Type III included six QTLs which were detectable only under salt stress, suggesting that these QTLs were apparently induced by the stress. Thirteen QTLs affecting trait difference or trait stability of ILs between the stress and non-stress conditions were identified and the Binam alleles at all loci except QPh4, QTn2 and QFw2a decreased trait difference. The three QTLs less influenced by the stress and 13 QTLs affecting trait stability were considered as ST QTLs which contributed to ST. Comparing the distribution of QTLs detected at the seedling and tillering stages, most (69%) of them were genetically independent. Only four were the same or adjacent regions on chromosomes 1, 2, 8 and 11 harboring ST QTLs detected at the two stages, suggesting that partial genetic overlap of ST across the two stages occurs. It is likely, therefore, to develop ST rice variety for both stages by pyramiding of ST QTLs of different stages or selection against the overlapping QTLs between the two stages via marker-assisted selection (MAS).