The present status of the atomic and molecular database required in fusion plasma edge modelling and diagnostics is briefly reviewed. Emphasis is placed on the evaluated and recommended data, and on the existing gaps and uncertainties in the database. Spectroscopic, collisional and surface data are considered.