The ability to age white-tailed deer (Odocoileus virginianus) is essential for population management, but field-aging techniques based on visual assessment of tooth-wear patterns lack accuracy. We used regression analysis to relate tooth measurements to age of 54 known-age wild South Texas (USA) male deer captured from 2002 to 2012. Using 9 animals/age class, from 2.5 to 7.5 years old, we measured cusp height and widths of each layer of enamel or dentine within each cusp of the third premolar (P3) on both left and right sides of the lower jaws. We found a linear relationship between age and dentine width (D), particularly within the anterior buccal cusp of P3 on the right jaw (r2 = 0.727). The regression equation, Age = 1.819 + (1.755 × D) provided a predictive aging model. We validated this model with 140 jaws from different known-age male deer within the same age range and from the same locations. Placement within the correct year class was achieved for 48% of male deer, and 90% were classified within 1 year of their actual age. This accuracy was greater than that achieved by 27 Wildlife Society members who correctly aged 28% of a subsample of the same jaws by tooth-wear patterns. Thus, using tooth measurements to age deer provided more accurate age estimation than visual tooth-wear methods. The dentine method is particularly useful for deer ≥3.5 years old. This technique may need to be calibrated with measurements from local known-aged deer before being used to age animals from regions beyond South Texas. © 2013 The Wildlife Society.