Despite multiple advantages of the use of electric shock as an aversive stimulus, reasons exist for considering alternative aversive stimuli. In the present study, we examined and compared the acquisition of two-way shuttlebox avoidance with 275.8-kPa (40-psi) pulsed air and continuous 0.4-mA shock in two strains of mice commonly employed in targeted gene mutation research, C57BL/6J and 129X1/SvJ. Each trial consisted of a 5-s warning stimulus (WS, light) during which shuttling to the other side cancelled delivery of the aversive stimulus. Once initiated, the aversive stimulus remained active for 20 s or until an escape response occurred. For C57BL/6J mice, air and shock were equally and highly effective aversive stimuli. In contrast, air was less effective than shock for 129X1/SvJ mice. C57BL/6J mice outperformed 129X1/SvJ mice for both stimulus types. For 129X1/SvJ mice, longer escape latencies were observed initially for air, suggesting that shock is more effective. However, these differences in latency dissipated within the first seven sessions. Nevertheless, by the end of the 17-day study, asymptotic levels of avoidance proficiency were substantially lower for air than for shock in 129X1/SvJ mice. These results indicate that air is a suitable substitute for shock as an aversive stimulus in shuttlebox active avoidance; however, the relative efficacies of these aversive stimuli appear to depend upon the strain chosen for study.