The presence of thin layers of free metal atoms at around 90 km in the upper atmosphere has been known for about fifty years. Layers of the alkali metals Na, K and Li, as well as Ca and Fe, have been observed. This discovery has posed two important questions. First, what is the source of the metals: interplanetary or terrestrial? Secondly, what is the nature of the chemistry that causes reactive metals such as sodium to exist in their atomic form in the atmosphere? The first part of this review covers the techniques that have been developed to observe the metal layers, including ground-, rocket- and space-based photometers, and in particular metal lidars. The many curious phenomena that have been observed are then described, such as the small scale-heights of the layers, the quite different seasonal variations of the three alkali metals, the large depletions of Ca and Fe relative to Na, and the dramatic appearance of sporadic Na layers. The second part of the review describes the recent advances that have been made in laboratory kinetic studies of metal reactions of atmospheric interest. A number of specialized techniques for making low-temperature measurements are compared, and a compilation of recommended rate coefficients is given. The history of modelling of the chemistry of metals in the mesosphere is then reviewed, and the evidence that their major source is meteoric ablation is presented. A current model of sodium is then described and evaluated with sensitivity tests. This section ends by discussing the possible impact of these meteoric metals on the chlorine-catalysed removal of ozone in the stratosphere. The final part of the review summarizes the active current state of the field and identifies some of its future directions.