As often happens in metrology, over ten years passed before metrologists had sufficient confidence in the new laser-based techniques for them to be considered a serious contender as a wavelength standard. Even when the technical case was made, the CIPM wanted to make sure that any change in the definition of the length unit could easily be accommodated by science, commerce, education, etc. It was also aware that there were several possible options for a redefinition. The two most obvious were to use a specific radiation – that of the 633 nm iodine-stabilized system, for example, as it was practical and convenient for much distance measurement. However, the CIPM was also attracted to a definition that was related directly to the fundamental constants and therefore preferred to wait until it had the results of several ambitious and difficult experiments that were designed to relate the length and frequency (time) standards in a measurement of the speed of light.

Complex "frequency chains" were set up at a number of national laboratories and resulted in international agreement on a number to be chosen for a fixed value of the speed of light with an uncertainty commensurate with the accuracy of the realization of the second and of the optical, laser based, value for the metre. The result was that the 1983 definition of the metre became based on the distance that light travels in free space in 1/299 792 458 of a second: 299 792 458 m s ^–1 being the agreed fixed value of the speed of light. This is still the current definition.

Several laser systems were possible contenders for a way of realizing this definition in practice. All employed a technique that came to be known as saturation absorption spectroscopy. The "basic" gas laser itself is essentially a discharge tube placed between two mirrors in a rigid cavity. The frequency (or wavelength) of these simple devices depended directly on the separation of the mirrors and so if the length of the laser cavity changed, so did its frequency. Lasers were also tuneable to a small extent because of the natural or Doppler width of the lasing transition. The idea behind saturated absorption spectroscopy was to place, inside the laser, a tube of a gas that had an absorption line that fell within the tuneable bandwidth of the laser. When the laser wavelength and the absorption wavelength were in coincidence a highly sophisticated servo system "locked" the laser wavelength to the centre of the absorption line. Since the conditions within the absorption cell could be carefully controlled it was possible, with a few extra tricks, to have absorption linewidths that were much less than the tuneable bandwidth of the laser. The popular laser systems that were developed included three which were shown to have a high stability: the red 633 nm line of a helium-neon laser in coincidence with an iodine absorption; the 3.39 µm helium-neon laser with methane; and the 10.6 µm laser with carbon dioxide. Eventually the red laser system was favoured because of its ease of operation and its immediate application to visible interferometry. The CCL now maintains a list of agreed wavelength (frequency) values for this and several other laser systems and discharge tube sources that can be used to realize the metre (the so-called Mise-en-Pratique).

There are, however, exciting and potentially highly influential developments in optical frequency standards, and the best optical ion and atom traps are now demonstrating a better frequency stability than the best stabilized lasers and the caesium-based microwave frequency standard. These systems use laser-cooling methods – a technique that won Bill Phillips, Steve Chu and Claude Cohen-Tannoudji the 1997 Nobel Prize for physics – to isolate and carry out experiments on a single isolated ion or atom. It is too early to speculate on how these may affect the definition of the metre – if at all – because for practical interferometry, spectroscopy and distance measurement, the current definition of the metre and its laser-based practical realization are perfectly adequate. However, there is another recent development – that of femtosecond laser-based frequency combs – which may well have a more immediate influence on the lasers used to realize the length standard.