Activity-triggered, rate-variable pacing systems offer a wide range of pacemaker patients the benefits of an appropriate paced Chronotropic response. However, optimizing settings for these devices often entails treadmill exercise testing. To assess simpler procedures for predicting appropriate settings, pacing rates of "strapped-on" and implanted Activitrax pacemakers were evaluated during arm exercise, walk-in-place exercise and treadmill exercise (0% slope at 1.5, 2.0 and 3.0 mph). For exercise of similar duration, steady-state pacing rates of implanted devices during arm and walk-in-place exercise were lower than those achieved during treadmill exercise. Linear regression analysis resulted in the slope of these relations most closely approaching 1.0 for arm exercise at "low" activity threshold, while walk-in-place tended to result in slopes approaching 1.0 at all activity thresholds. Similarly, although strapped-on devices underestimated rates obtained with implanted devices, differences fell within a narrow range (-6 to -14%). Thus, in patients with implanted or strapped-on Activitrax pacemakers, relatively simple exercise procedures suitable for office or clinic environment may be useful to provide an estimate of pacing rates during physical activity, and may thereby facilitate selection of appropriate programmable settings.