or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. conservation practitioners and managers advocate the use of LHS (Jones, 2003;McKay et al., 2005), which theoretically is adapted to the local environmental conditions of a region (McMillan, 1969) and maintains the genetic integrity of local plant populations (Bischoff et al., 2010). However, LHS may not be suited for agronomic applications such as repeated defoliations under rotational grazing and as a result may not ABSTRACT Native warm-season grasses have potential to increase the ecosystem services associated with grazing lands and to provide a source of forage for livestock. In the upper Midwest, minimal region specifi c information is available to grass-based farmers on how to manage native pastures for long-term productivity and persistence. Wildlife biologists suggest delaying defoliations until mid-summer to preserve grassland bird habitat, yet production-based management recommendations warn against late season defoliations. Moreover, the source of seed to use for pasture establishment, cultivar or locally harvested seed (LHS), is unclear. We initiated an on-farm experiment to evaluate the eff ects of two rotational grazing schedules, June to July or July to September, for diff erences in vegetation cover, vigor, yield, and nutritive value of paddocks seeded with native warm-season grass cultivars or LHS. Aft er 2 yr, native grass cover was similar across all treatments. Vigor, as measured by spring etiolated growth, was sustained under grazing and across seed sources for the dominant native grasses big bluestem (Andropogon gerardii Vitman) and Indiangrass [Sorghastrum nutans (L.) Nash]. Total yield was greatest when grazing was deferred until July, but this resulted in decreased nutritive value for both seed sources. Th ese results indicate that native warm-season grasses have potential to persist under rotational grazing in upper Midwest pasture systems and that pasture performance is controlled by the timing of the fi rst grazing cycle. Th e lack of a consistent seed source eff ect suggests that the LHS and cultivars used in the study are suitable for pasture use in the region.