The rate constants of O- + N2 → N2O + e- from 800 K to 1200 K and the reverse process e- + N2O → O- + N2 from 700 K to 1300 K are measured using a flowing afterglow - Langmuir probe apparatus. The rate constants for O- + N2 are well described by 3 × 10-12 e-0.28 eV kT-1 cm3 s-1. The rate constants for e- + N2O are somewhat larger than previously reported and are well described by 7 × 10-7 e-0.48 eV kT-1 cm3 s-1. The resulting equilibrium constants differ from those calculated using the fundamental thermodynamics by factors of 2-3, likely due to significantly non-thermal product distributions in one or both reactions. The potential surfaces of N2O and N2O- are calculated at the CCSD(T) level. The minimum energy crossing point is identified 0.53 eV above the N2O minimum, similar to the activation energy for the electron attachment to N2O. A barrier between N2O- and O- + N2 is also identified with a transition state at a similar energy of 0.52 eV. The activation energy of O- + N2 is similar to one vibrational quantum of N2. The calculated potential surface supports the notion that vibrational excitation will enhance reaction above the same energy in translation, and vibrational-state specific rate constants are derived from the data. The O- + N2 rate constants are much smaller than literature values measured in a drift tube apparatus, supporting the contention that those values were overestimated due to the presence of vibrationally excited N2. The result impacts the modeling of transient luminous events in the mesosphere.