The theory of high-temperature superconductivity in oxides is discussed, proposing that charge carriers are holes in the O(2p) states and pairing is mediated by strong coupling to local spin configurations on Cu sites. High-temperature superconductors have transition temperatures much higher than those explained by phonon exchange, suggesting an alternative mechanism involving strong coupling to spin configurations. This mechanism is similar to anisotropic pairing via spin fluctuations but is stronger due to the electronic and crystal structure of oxides. The model differs from Anderson's resonant-valence-bond model and has an antiferromagnetic insulating limit. The high-temperature superconductors have quasi-two-dimensional electron motion in CuO₂ planes. The Hamiltonian for a single plane is given by an extended Hubbard model, with parameters such as t, ε, U_p, U_d, and V. The model is applied to La₂CuO₄, where the number of holes per Cu site is 1 ± δ. Electronic band structure calculations suggest that La₂CuO₄ has a half-filled band, but the presence of La(5d) band dipping below the Fermi level removes electrons from CuO₂ planes, leading to δ > 0. Doping with Sr or Ba removes electrons from the La(5d) band, increasing δ. For t=0 and δ=0, the ground state has one hole per Cu site. Hopping leads to an effective Hubbard model with hopping integral t_d = t²/(ε + V). Monte Carlo studies show that the ground state is an antiferromagnetic insulator, and long-range antiferromagnetic order exists at finite temperatures. A gap Δ exists between occupied and unoccupied states, and additional holes go into O(2p) states if the site energy lies within the gap. An effective Hamiltonian for O(2p) holes is derived, showing that the charge carriers have a narrow band -t̄e_k and a number density n_c = nδ. The effective attractive coupling between O(2p) holes is responsible for superconductivity, with the coupling being strong due to O(2p)-Cu(3d) exchange interactions being larger than Cu(3d)-Cu(3d) exchange. The BCS transition temperature is determined by the condition for a nontrivial solution of the gap equation, leading to a d-state pairing with T_c ~ E₀e^{-7πt̄/v₀}. The model predicts T_c between 30 and 40 K, with E₀/k_BT_c ~ 20. As δ increases, T_c decreases. The high T_c in YBa₂Cu₃O₉-δ may be due to increased carrier density. The model suggests that real pairs, not Cooper pairs, may form for small δ. The theory is consistent with experimental observations and provides a framework for understandingThe theory of high-temperature superconductivity in oxides is discussed, proposing that charge carriers are holes in the O(2p) states and pairing is mediated by strong coupling to local spin configurations on Cu sites. High-temperature superconductors have transition temperatures much higher than those explained by phonon exchange, suggesting an alternative mechanism involving strong coupling to spin configurations. This mechanism is similar to anisotropic pairing via spin fluctuations but is stronger due to the electronic and crystal structure of oxides. The model differs from Anderson's resonant-valence-bond model and has an antiferromagnetic insulating limit. The high-temperature superconductors have quasi-two-dimensional electron motion in CuO₂ planes. The Hamiltonian for a single plane is given by an extended Hubbard model, with parameters such as t, ε, U_p, U_d, and V. The model is applied to La₂CuO₄, where the number of holes per Cu site is 1 ± δ. Electronic band structure calculations suggest that La₂CuO₄ has a half-filled band, but the presence of La(5d) band dipping below the Fermi level removes electrons from CuO₂ planes, leading to δ > 0. Doping with Sr or Ba removes electrons from the La(5d) band, increasing δ. For t=0 and δ=0, the ground state has one hole per Cu site. Hopping leads to an effective Hubbard model with hopping integral t_d = t²/(ε + V). Monte Carlo studies show that the ground state is an antiferromagnetic insulator, and long-range antiferromagnetic order exists at finite temperatures. A gap Δ exists between occupied and unoccupied states, and additional holes go into O(2p) states if the site energy lies within the gap. An effective Hamiltonian for O(2p) holes is derived, showing that the charge carriers have a narrow band -t̄e_k and a number density n_c = nδ. The effective attractive coupling between O(2p) holes is responsible for superconductivity, with the coupling being strong due to O(2p)-Cu(3d) exchange interactions being larger than Cu(3d)-Cu(3d) exchange. The BCS transition temperature is determined by the condition for a nontrivial solution of the gap equation, leading to a d-state pairing with T_c ~ E₀e^{-7πt̄/v₀}. The model predicts T_c between 30 and 40 K, with E₀/k_BT_c ~ 20. As δ increases, T_c decreases. The high T_c in YBa₂Cu₃O₉-δ may be due to increased carrier density. The model suggests that real pairs, not Cooper pairs, may form for small δ. The theory is consistent with experimental observations and provides a framework for understanding