Light-emitting diodes (LEDs) are widely used as artificial smart light sources in lighting, display, communications, medical services, and signage. Recently, organic-inorganic hybrid lead halide perovskites LEDs have drawn vigorous scientific interest due to their low cost, high efficiency, and simple manufacturing process. Up to now, the external quantum efficiency (EQE) of green, red and near-infrared perovskite LEDs have been improved to more than 20%. Despite these exciting progresses, the toxicity of lead would significantly hinder their further development and commercialization. Extensive research effort has been devoted to the development of lead-free halide perovskites. The most attractive candidate is double perovskites (A2B+B3+X6), where divalent Pb2+ is replaced by monovalent B+ and trivalent B3+ metal ions, in terms of similar crystal structure and a wide range of possible combinations. More importantly, lead-free halide double perovskites usually exhibit broadband and large Stokes shift emission derived from self-trapped excitons (STEs), which is very attractive for white light-emitting diodes (WLED) with single emissive layer. Unfortunately, the device efficiencies are still extremely low. The underlying reasons possibly relate to indirect bandgap and parity-forbidden transitions between bandgaps, extremely low photoluminescence quantum yield (PLQY), non-uniform films with defects, and ineffective carrier injection process. In this project, we aim to address the low efficiency of lead-free perovskite WLED devices through the following strategies. Firstly, we will develop novel and stable 3D lead-free double perovskites with direct bandgap through a combination of first-principles calculation and the Goldschmidt tolerance/octahedral factor. Meanwhile, B-site metal doping/alloying strategy will be employed to break the parity-forbidden transitions between bandgaps. Secondly, we will enhance the photoluminescence properties of these new double perovskites by introducing suitable A-site cation to reduce the 3D crystal structure to a 2D layered structure. We expect that the low dimensional crystal structure could facilitate octahedral distortion and the formation of STEs, as STEs is a kind of exciton luminescence and is highly related to the Jahn-Teller distortion in octahedra. We will further facilitate the emission of STEs by forming the quantum well structure between organic layer and inorganic octahedrons layer in 2D double perovskites to increase the exciton binding energy. Thirdly, we will find an appropriate preparation process to fabricate high quality 2D double perovskite films, such as spin coating with antisolvent engineering or thermal evaporation. Finally, we will analyze and model the performance limiting processes in these WLED devices through device physics to further screen out a suitable electron/hole transport layer (ETL/HTL) and guide the optimization of device efficiency. Based on the above strategies, we aim to achieve operational lead-free perovskite WLED.