It has been suggested that a developmental log-to-linear shift in childrens performance on number line estimation tasks is diagnostic of their underlying representations of numerical magnitude (Siegler & Opfer, 2003). However, in the study presented herein, we were able to induce a similar log-to-linear shift on number line estimation tasks among adults by manipulating their familiarity with the numbers we used as stimuli. We offer this evidence as an existence proof that differences in performance on number line estimation tasks may not necessarily be indicative of fundamental differences in the formats of peoples underlying numerical magnitude representations. Rather, they may be diagnostic of differences in peoples understandings of what magnitudes are represented by symbolic numbers.