A cereal grain harvesting system is introduced that combines existing technologies in a unique way to improve cereal grain harvest performance, increase profitability, and efficiently collect biomass. The harvesting system is comprised of three machines: one to reap grain, harvest biomass, and size crop residue for no-till seeding; a second to thresh and winnow the grain; and a third to separate the grain by quality for added value. This study describes the new harvesting system and the development of one of the system's major components: the reaper/flail harvester. The reaper/flail harvester consists of a mobile power unit, a stripper header to harvest the crop, and a flail to chop the standing residue into small pieces. A prototype harvester was fabricated and tested to determine system design criteria and performance characteristics in terms of machine power requirements, quantity of biomass collected, and bulk density of the material harvested. Trials were conducted in seven wheat (Triticum aestivum L.) fields in Oregon during 2005 and 2006 that ranged in yield from 3.3 to 6.4 t ha-1. Harvester performance was evaluated at various travel speeds, straw chop heights, and with different types of wheat. Flail power requirements were highly linearly correlated with quantity and rate of biomass chopped (R 2 = 0.91). The maximum reaping power requirement was 2.7 kW m-1 , only slightly higher than the no-load power requirement of 1.9 kW m-1. Power requirements for reaping, conveying, and flailing ranged from a low of 5.0 to a high of 13.5 kW m-1 depending on travel speed, crop yield, biomass concentration, and chop height. Values were linearly correlated with the combined grain, chaff, and biomass feed rate (t h-1) with an R 2 of 0.88. Total machine power requirements for a harvester with a 7.3 m header would be about 175 kW, including 75 kW for propulsion, losses, and reserve. Chaff yield in the grain/chaff (graff) mixture harvested exceeded 2 t ha-1 in six of the seven trials. With chaff valued at $23 t-1 , collecting 2 t ha-1 of chaff would increase farm revenues by $46 ha-1. Realistic graff densities of awned wheat were less than 1/11 that of clean grain, and new, efficient material handling systems would need to be developed to have harvesting capacities comparable to that of a conventional combine-based system. Awnless wheat had graff densities that averaged about 1/5 that of clean grain. Equipment is commercially available to handle this volume of material and have harvesting field capacities comparable to that of a conventional combine-based system.