According to the pilot, he topped off both wing fuel tanks about 2 weeks prior to the accident flight. The engine was ground run for about 10 minutes on the day after the fueling, after which the airplane was parked until the day of the accident flight. During the preflight inspection for the accident flight, the pilot sampled fuel from the fuel drains, but did not visually inspect the level of fuel in the fuel tanks, and assumed they were both full. About 1 hour into the flight, the engine started to “miss”, and the pilot then activated the fuel boost pump. The engine then ran normally for about 15 to 20 seconds before it began to “miss” again, until a few moments later when it lost all power. The pilot attempted to divert to a nearby airport, however during the approach he determined he would not reach the runway, and he selected a field for a forced landing. During landing the airplane struck trees, and a portion of the left wing separated about 3 ft outboard of the root. The airplane impacted terrain, came to rest inverted, and sustained additional substantial damage to the right wing, forward fuselage, horizontal stabilizer, and elevator. A Federal Aviation Administration inspector observed that both of the left wing fuel tanks were breached and that only a trace amount of fuel was found in the fractured left wing fuel line near the fuselage. The right wing fuel tanks remained intact but were inaccessible. The inspector reported, and the pilot confirmed, that there was no odor of fuel at the accident site. Following the accident, pilot reported that there were no preimpact mechanical malfunctions or failures of the airplane that would have precluded normal operation prior to the accident. He concluded that although he believed the fuel tanks were full prior to departure, they likely were not, and he did not confirm the fuel level during the preflight inspection. The pilot further described that the airplane was equipped with fuel quantity gauges, however the gauges were located in the footwell for the rear seat, near but below his hips when seated in the front seat, and they were difficult to see from the front seat position. He added that the fuel quantity gauges were of limited use. Due to the wing dihedral, the gauges would indicate “full” until about 8-12 gallons had been used, and then read empty after using another 3-4 gallons. He had instead used the fuel computer/totalizer indications to reference fuel quantity. He had previously calibrated the totalizer by comparing its indications with fueling records several times and found it to be accurate. Given this information it is likely that the loss of engine power was the result of fuel exhaustion.