In the Marsyandi valley of central Nepal, a major strand of the South Tibetan detachment system, the 18--22 Ma Chame detachment, places epidote-amphibolite to amphibolite facies calc-silicate rocks and marbles of the Tibetan sedimentary sequence on amphibolite facies pelitic gneisses and calc-silicate rocks of the Greater Himalayan sequence. Although the resulting metamorphic discontinuity is minor and sometimes cryptic, 40Ar/39Ar thermochronologic results from the area reveal that the hanging wall and footwall of the detachment had distinctive thermal histories. Hanging wall phlogopites and biotites yield cooling ages of 27.1--29.9 Ma, compared with footwall biotite ages of 14.1--16.6 Ma. U-Pb monazite thermochronology demonstrates that the Greater Himalayan sequence experienced peak amphibolite facies conditions at approximately 22 Ma, but the 40Ar/39Ar results require hanging wall metamorphism to be an Oligocene (or older) phenomenon. These events are interpreted as representing the Neohimalayan and Eohimalayan metamorphic phases proposed by previous workers in the central Himalaya. Some of the dated hanging wall phlogopites grew synchronously with development of SW vergent macroscopic folds in the Tibetan sedimentary sequence, implying that Eohimalayan metamorphism was associated with an important phase of crustal shortening in this sector of the Himalaya. Despite the intensity of Neohimalayan metamorphism below the Chame detachment, evidence for Eohimalayan metamorphism and igneous activity is preserved in the footwall rocks of the Marsyandi drainage. Inherited approximately 35 Ma monazites of either metamorphic or igneous origin have been found in the upper Greater Himalayan sequence in this area, and one hornblende separate from the uppermost footwall yields a 40Ar/39Ar age of 30.0¿3.0 Ma. This hornblende date and a similar result from the same structural level in the nearby Kali Gandaki valley, if robust, suggest that the duration of the Neohimalayan event was extremely short, probably less than 1 million years and that peak temperatures in the upper part of the Greater Himalayan sequence were not substantially greater than about 900 K. ¿ 1998 American Geophysical Union