Previous panel research has provided individual evidence for aspects of the U-shaped pattern, but these studies typically rely on sampling the same speaker at two points in time, usually in close proximity. As a result, our knowledge about the patterning of age-graded variables across the entire adult life-span is limited. What is needed, thus, is a data-set that captures ongoing linguistic malleability in the individual speaker across all “life experiences that give age meaning” (Eckert 1997:167). Our study is the first to add real time evidence across the lifespan as a whole on an age-graded variable. We present the results of a novel dynamic data-set that allows us to model speakers’ linguistic choices between ages 19 and 78. We illustrate the age-graded patterns in our data and draw attention to the complex, socially niched ways in which speakers react to age-specific expectations.