On July 25, 1997, at 1909 hours mountain standard time, a Classic Aircraft Waco YMF, N369BP, collided with a signpost during the ground roll portion of an emergency landing on a roadway near Sedona, Arizona. The forced landing was precipitated by a partial loss of engine power during cruise. The aircraft sustained substantial damage, and the pilot and two passengers were not injured. The aircraft was operated by Red Rocks Bi-Plane Tours of Sedona as a for-hire sightseeing flight under 14 CFR Part 91 when the accident occurred. The flight originated in Sedona at 1845 on the day of the accident. Visual meteorological conditions prevailed and no flight plan was filed. The pilot stated that approximately 10 minutes into the flight he noticed roughness in the engine and turned back towards the airport. Shortly thereafter he noticed a decrease in power, which continued to escalate. When the power reached a point that he was unable to maintain altitude, he decided to attempt to find a suitable landing site and landed the aircraft on the centerline of a local roadway. He stated that approximately 100 feet into the rollout, about 60 mph, the right wing struck a telephone junction box. This caused the airplane to veer to the right. The airplane and engine were examined under the supervision of the Safety Board on July 31, 1997. Control continuity was established for the throttle, carburetor heat, and the mixture controls. The air intake screen was clean with no debris found. Both magnetos were examined and rotated, with sparks developed at each ignition lead. The number 6 cylinder exhaust valve rocker arm adjustment nut was found loose, and the valve would not operate with crankshaft rotation. Jacobs Engines, the manufacturer of the engine, was contacted for information on the effect of a nonoperational exhaust valve on the engine's ability to produce power. The representative reported that a nonoperational valve could effect power output; however, they had no test data to quantify the effects. A request was made to the company to perform a test cell run in an attempt to duplicate the problem and quantify the amount of power loss. Impact damage precluded use of the accident engine for the test. As of the submission of this report, the manufacturer had not been able to locate an engine to run on the test cell. Review of the aircraft maintenance records disclosed that the engine was returned to the factory on February 18, 1997, about 200 hours prior to the accident, for repair of a broken internal cluster gear. The logbook notes in that entry that the valves were readjusted at that time. Examination of all subsequent entries disclosed no record of valve adjustments. The aircraft and all associated components were turned over to the insurance company representing the owner on July 31, 1997.