This paper examines social influences on the realization of voiced stops in inland California. We analyzed sociolinguistic interviews with 62 white residents from Redding, Merced, and Bakersfield (which mark the northern, middle, and southern points of California’s Central Valley), balanced for sex, class, age, and whether a speaker earns their livelihood off the land. We follow Jaciewicz, Fox, and Lyle (2009) in examining the extent of voicing during stop closures (duration of voicing during closure relative to total duration of closure), and also adopt a novel measure of the magnitude of voicing, which captures the intensity of a stop closure relative to the following vowel. Mixed effects linear regression models were constructed for both voicing measures, with a number of linguistic and social predictors considered in addition to random effects. Results show that the extent of voicing measure was insufficiently sensitive to differentiate speakers, as nearly everyone exhibited voicing throughout the closure. The voicing intensity measure, however, was shown to reveal significant effects of place of articulation, closure duration, and ties to the land. Most importantly, speakers who earn their livelihood off the land exhibit significantly stronger voiced stops than those who do not. We argue that even though strongly voiced stops likely entered California during a large-scale in-migration of Southerners during the Dust Bowl (Jaciewicz et al. 2009 report more extensive voicing among women from the South compared to the Midwest), they have since taken on locally significant indexicalities reflecting the values and ideals of land-oriented communities throughout the Central Valley (and do not simply mean “Southern”). Our findings also raise questions about where the linguistic limits of socially structured variation lie, given the systematic social patterning observed here for low-level phonetic details (i.e., voicing intensity) that likely operate far below the level of consciousness.