Designed by Rex Beisel, the XO-l prototype was tested with a Liberty engine by Lt. Harold Harris on November 30, 1924, and showed “remarkably high performance,” excellent flying qualities, and “handled exactly like a pursuit plane.” Doubts were raised by the observation board about its suitability for “average pilots,” and the maintenance of its riveted duralumin tube fuselage structure. The board suggested a service test purchase of the Curtiss type.
On December 1, 1924, the XO-l, like the XO-2, was also tested with the Packard lA-1500, and although the Curtiss competitor’s performance was best, no order for that version was placed. The Wright XO-3 tested on January 23, 1925, was a double-bay biplane with an experimental Wright V-1950 engine. It was not purchased by the Army, and the Wright company abandoned efforts at aircraft construction to concentrate quite successfully on engines. Another observation project of the period, a Liberty-powered Martin XO-4, was canceled before completion.
Ten Curtiss Falcons were ordered for service test, the first appearing in February 1926 with the small and neatly cowled Curtiss D-12 (V-1150), while the ninth was tested in October 1926 as the O-lA with a 430-hp Liberty, and another became the XO-11. That Liberty version was first ordered in 1926 for observation squadrons as the O-11. Sixty-six O-11s went into service in 1927-28, armed with the one .30-caliber fixed and twin .30-caliber flexible guns usual on observation planes. They served with the 5th, 12th and 99th Squadrons, as well as with the National Guard.
As for the Army’s attack units, a board convened on August 2, 1926 had agreed that the Curtiss was the most suitable model until a more satisfactory type could be developed. On February 2, 1927, 65 Falcons were ordered, and a September contract added 36 more. They were delivered as 76 A-3, 21 O-lB, and four O-lC Falcons, the latter being unarmed official transports.
The first A-3 was ready by October 31, 1927. A 435-hp Curtiss D-12D engine was under the smooth cowling, the radiator was underneath, and a neat prop spinner gave the nose the characteristic “Eversharp pencil” appearance. The wood-frame wings used the Clark Y airfoil, with the upper wing swept back from a straight center section, while the lower wings were straight. There were two M1921 Browning guns in the nose, two wing guns firing forward, and two flexible Lewis rear guns. All were .30-caliber, with 400 rounds provided for the wing guns and 600 rounds for the others. Wing racks for 200 pounds of fragmentation bombs were attached, or a 56-gallon auxiliary tank could be carried behind the tunnel radiator.
Weapons were removed and dual controls installed on six A-3s, which became the A-3A trainers. The second A-3 was completed with the new air-cooled Pratt & Whitney Wasp radial and tested in May 1928 as the XA-4. Twenty-one O-lB observation versions were similar to the A-3, but had only one Browning gun firing forward.