A record of an investigation of heavy-duty truck-axles carried out by the Bureau of Standards at the request of the motor transport division of the Quartermaster Corps, this paper deals in particular with the mechanical efficiency of the axles tested. The investigation included, in addition to several worm-drive Class-B Army-truck axles with different types of bearing, a number of axles usually designated respectively as “internal-gear” and “multiple-reduction” type. Each of these types was represented by axles in successful use in commercial trucks of 5-ton rating.In analyzing the results of the tests it was found possible to separate the losses into no-load losses and load losses; the total loss being the sum of these two. In general, the no-load losses were primarily controlled by the viscosity and the method of application of the lubricant. They were greater in those axles in which the parts rotating at high speeds were immersed most completely in the lubricant. Where a separate body of lubricant was used in the wheels, the losses were increased still further. In a given axle, these losses also increased with an increase in the viscosity of the lubricant and with the propeller-shaft speeds.The load losses, expressed as torque, were practically independent of the speed and of the viscosity and the method of application of the lubricant. They increased, however, with an increase of the torque input and at an increasing rate. In the case of the several Class-B worm-type axles tested, the increase in the load losses with an increased torque-input was greater than with any of the gear-type axles.The use of a lubricant of lower viscosity at low temperatures would improve the mechanical efficiency under the more unfavorable conditions of speed and temperature, and seems advisable. Such lubricant should, however, be no less efficient than that used in the tests at the higher temperatures and the heavy pressures resulting from high torque-values. Improvement in the method of circulating the lubricant is suggested as desirable by the results; the high no-load losses of some of these axles show the present method of circulation to be very inefficient. The characteristic difference between the worm-drive and gear-drive types is demonstrated by the greater load losses of the former type.