A method and apparatus for inter-layer prediction for scalable video coding are disclosed. Embodiments of the present invention utilize weighted prediction for scalable coding. The weighted prediction is based on the predicted texture data and the inter-layer Intra prediction data derived from BL reconstructed data. The inter-layer Intra prediction data corresponds to the BL reconstructed data or up-sampled BL reconstructed data. The predicted texture data corresponds to spatial Intra prediction data or motion-compensated prediction data based on the second EL video data in the same layer as the current EL picture. Embodiments of the present invention also utilize the reference picture list including an inter-layer reference picture (ILRP) corresponding to BL reconstructed texture frame or up-sampled BL reconstructed texture frame for Inter prediction of EL video data. The motion vector is limited to a range around (0,0) when the ILRP is selected as a reference picture.