A feeding-trial procedure was employed as a means for estimating the net energy value of feedstuffs for milk production, using an adaptation of the feed increment principle. The reported observations were made over a period of three years in four trials with Latin-square designs. Basal rations made up of 70 per cent alfalfa and 30 per cent barley were tested against increments of alfalfa hay, barley, dried beet pulp, dried beet pulp with molasses, dried beet pulp with concentrated Steffen’s filtrate, and milo.

Results indicated that the procedure could be useful for routine evaluation of feeds. Repeatability appeared encouraging as shown by net energy values of 71.8, 68.4 and 68.8 mcal/100 lb. dry matter, determined with three different lots of barley in as many years.

The main problem yet to be solved is the determination of milk energy equivalent of liveweight change. In this regard the greatest source of error was the estimating of liveweight change.