The degree to which desertification during the Holocene resulted from climatic deterioration or alternatively from overgrazing has puzzled Quaternary scientists in many arid regions of the world. In the research reported upon here, a multi-disciplinary investigation of a 5-m deep, ~11,000-year-old sediment column excavated in a dry lake bed in southern Jordan revealed an early interval in which proxies of plant cover and sheep/goat stocking rates co-varied directly with climatic cycles. Beginning ~5.6 kcal BP, however, this pattern changed with fecal spherulite and phytolith densities failing to co-vary and with spherulites often greatly exceeding phytolith densities, which we suggest is indicative of overgrazing. Moreover, the lack of agreement between the high phytolith densities and other indicators of a desert landscape (i.e. geomorphic and palynologic) suggests that phytolith densities were inflated by fodder subsidies and as such are not entirely reflective of plant cover for this later interval. Given the co-incidental emergence of overgrazing with archaeological evidence for a marked rise in regional population, emergence of widespread trade, and expansion of the Timnian pastoral complex during Chalcolithic–early Bronze times, we argue that desertification was a consequence of socio-economic factors (e.g. higher stocking rates) associated with a shift from a subsistence to a market economy. In addition, we contend that the signature lithic artifact variety (tabular scraper) that appeared in great abundance during this period was directly tied to the emergent market economy and its secondary products (wool) in having been used to shear sheep. Moreover, in that these changes took place largely concurrent with local and regionally recognized evidence of a moist interval, we conclude that the mid- to late-Holocene desertification of the southern Levant was induced more by anthropogenic than climatic factors.