The 40Ar/39Ar dating of postcollapse rhyolites in the Valles Caldera (1140 kyr) defines five intervals during which rhyolite domes, flows, and pyroclastic rocks were emplaced from ring fracture vents. Previously, ages were based on limited K-Ar dating and stratigraphic relations. The 40Ar/39Ar dates are concordant with the previous dates for about half of the samples; others have ages significantly older (7--25%) despite evidence for small amounts of xenocrystic material. Eruptive activity occurred during relatively brief periods at 1133, 973--915, 800--787, 557--521, and ~300--170 kyr separated by quiescent intervals of ~115--300 kyr. Isochrons defined by single-crystal sanidine fusions indicate trapped 40Ar/36Ar ranging from ~282 to 326. Sanidine analyses yield reliable and consistent results, whereas biotite and glass data may be ambiguous. Biotite may have low radiogenic yields, contain variable amounts of extraneous argon, or define apparent isochrons discordant with cogenetic sanidine. Unaltered, unhydrated glass in one case yields an isochron concordant with coexisting sanidine, whereas in another is undateable due to large amounts of non-radiogenic argon. The revised chronology has implications for the evolution of the Jemez magmatic system as well as for the geomagnetic polarity timescale. The discrete periods of volcanism, relations among rhyolites erupted within these intervals, and the duration of the intervening quiescent intervals are consistent with the generation, evolution, and eruption of several distinct magmas following caldera collapse. In addition, the 40Ar/39Ar data suggest revision of the Brunhes--Matuyama polarity reversal boundary from the currently accepted 730 kyr to 780 kyr. ¿ American Geophysical Union 1993