The Multiple Mirror Telescope (MMT) has been used in experiments with sodium laser guide stars to sense and correct atmospheric image distortion. The major goal was to develop and test concepts for a full adaptative optics system, to be used in 2 yr when the present array of six 1.8 m telescopes will be replaced with a single 6.5 m mirror. The guide star, produced by a continuous-wave dye laser beam projected out along the optical axis of the telescope, was as bright in the V band as a natural star of m<SUB>v</SUB> = 10.4. Our tests culminated in the first demonstration of a sodium laser guide star used to improve the image of an astronomical telescope, in this case formed by two of the six 1.8 m apertures. Two adaptive servo loops were closed simultaneously. The laser beacon provided a measure of the differential wave-front tilt between the two apertures, and a natural guide star was used to measure the overall wave-front tilt. A factor of 2 improvement in the K-band Strehl ratio was measured, and the resolution improved from 0.58 arcsec to 0.41 arcsec. The experiment demonstrated all the features needed for correction of the 6.5 m telescope to the diffraction limit using a sodium beacon. The accuracy with which the laser beacon measures the atmospheric aberration of starlight across the full 6.9 m aperture of the MMT was examined. From more extensive measurements of binary star wave fronts, we deduce that focus anisoplanatism for the 6.5 m telescope will correspond to a Strehl ratio of typically 88% at K under normal seeing conditions. In a laser-based adaptive system, a natural guide star is still required to sense overall wave-front slope. Our measurements of binary stars also yielded the image degradation to be expected from differences in the overall slopes between the wave fronts from the object of scientific interest and the natural guide star. A Strehl ratio of 80% at K was deduced for an offset of 40 arcsec, implying that good sky coverage will be possible. In general, our results are consistent with calculations based on measurements of atmospheric turbulence at the best sites. Our direct measurements over such a large aperture show clearly the effects of a finite and variable outer scale of turbulence.