Density functional theory (DFT) has become a widely used tool in chemistry and materials science due to its low computational cost and useful accuracy. However, it faces several limitations, including the need for numerous approximations, poor performance for strongly correlated systems, and slow performance for liquids. This perspective reviews recent progress and ongoing challenges in DFT. Over the past 20 years, DFT has become a standard technique in various branches of chemistry and materials science. It is now applied to a wide range of problems, from predicting new catalysts and Li battery materials to studying molecular conductance and protein folding. The number of papers citing DFT has surged, reflecting its growing importance. Key developments include the introduction of meta-GGA functionals, which improve the accuracy of ground-state energies for molecules, solids, and surfaces. The HSE functional, which combines Hartree-Fock exchange with DFT, has shown promise in accurately predicting fundamental gaps in semiconductors and insulators. Additionally, the development of orbital-free DFT and partition DFT aims to address the limitations of traditional DFT by avoiding the need to solve the KS equations directly. Despite these advancements, DFT still faces challenges, such as the inability to accurately predict the fundamental gaps of semiconductors and insulators, and the difficulty in treating strongly correlated systems. The field is also exploring new paradigms, such as optimized effective potential (OEP) and random phase approximation (RPA), which offer potential improvements in accuracy and efficiency. The future of DFT lies in the development of more accurate and efficient methods, particularly in addressing the limitations of current approximations and expanding their applicability to larger systems.Density functional theory (DFT) has become a widely used tool in chemistry and materials science due to its low computational cost and useful accuracy. However, it faces several limitations, including the need for numerous approximations, poor performance for strongly correlated systems, and slow performance for liquids. This perspective reviews recent progress and ongoing challenges in DFT. Over the past 20 years, DFT has become a standard technique in various branches of chemistry and materials science. It is now applied to a wide range of problems, from predicting new catalysts and Li battery materials to studying molecular conductance and protein folding. The number of papers citing DFT has surged, reflecting its growing importance. Key developments include the introduction of meta-GGA functionals, which improve the accuracy of ground-state energies for molecules, solids, and surfaces. The HSE functional, which combines Hartree-Fock exchange with DFT, has shown promise in accurately predicting fundamental gaps in semiconductors and insulators. Additionally, the development of orbital-free DFT and partition DFT aims to address the limitations of traditional DFT by avoiding the need to solve the KS equations directly. Despite these advancements, DFT still faces challenges, such as the inability to accurately predict the fundamental gaps of semiconductors and insulators, and the difficulty in treating strongly correlated systems. The field is also exploring new paradigms, such as optimized effective potential (OEP) and random phase approximation (RPA), which offer potential improvements in accuracy and efficiency. The future of DFT lies in the development of more accurate and efficient methods, particularly in addressing the limitations of current approximations and expanding their applicability to larger systems.