The aircraft had been operating normally except for a report that the airspeed indicator was under-reading in cruise and the engine had been difficult to start for a number of days. On the morning of the accident the aircraft was flown from Karratha to Nanutarra in support of another aircraft. The pilot had been told that the Nanutarra strip was in good condition. However, this proved not to be the case as the surface was covered with brush up to one metre in height and the landing area was only ten metres wide. The landing was uneventful, but the pilot had to remove some brush from the landing gear after landing. Prior to departure on the accident flight the pilot discussed the condition of the strip with the pilot of the other aircraft. This second pilot, who owned VH-WMV, paced out 500 metres of reasonable strip and indicated the pilot should use this for takeoff. The pilot, still concerned, paced out a further 150 metres giving 650 metres in all. Although neither pilot referred to it, the takeoff distance chart in the Pilots Operating Handbook (POH) indicated that 426 metres of ground roll was required in the prevailing conditions. The pilot was also concerned about a shed at the end of the strip and decided that, in the event of an emergency during the takeoff, he would steer the aircraft to the left of the shed. The aircraft was again difficult to start and the owners assistance was needed. During the taxy for takeoff, the pilot had to use the auxiliary fuel pump, on several occasions, to keep the engine running. Although he completed the pretakeoff checks the pilot did not complete a magneto function check as he believed the strip surface was inadequate and he wanted to keep the use of high power settings to a minimum. The pilot did run the engine up to full power, against the brakes, before commencing the takeoff roll. At that point all engine indications appeared normal. During the takeoff roll the pilot concentrated his attention on keeping the aircraft in the centre of the very narrow section of reasonable strip. At the end of the 650 metres he selected the normal climb attitude and lifted off at, what he believed was, about 75 knots. Immediately after liftoff the stall warning sounded and the pilot lowered the nose slightly, at the same time turning left to avoid the shed. The aircraft continued to descend until it collided with trees. The aircraft came to a stop 150 metres to the left of the strip centreline and 1400 metres from the start of the takeoff roll. The pilot reported that after liftoff he felt that the aircraft was not performing to expectations and that the engine had lost power. A post accident inspection of the engine disclosed that the compression on two cylinders was probably below serviceable limits at the time of the takeoff. Also, the bottom plugs in two other cylinders may have been contaminated and not been working efficiently. The engine had been serviceable during the periodic inspection completed 87 hours prior to the accident. The reports that the airspeed indicator under-read in cruise, the engine was difficult to start and the difficulty encountered by the pilot in keeping the engine running during taxy may also have been indications of poor engine performance. The lack of normal acceleration, during the takeoff roll, was probably the result of the less than optimum performance from the engine. The drag caused by the brush on the runway may have also been a factor. The pilot was forced to liftoff at the 650 metre mark as the condition of strip, after that point, was poor. Although the pilot reported that liftoff speed was 75 knots, the onset of a stall immediately after liftoff indicates that the speed was below this figure. The POH gives a power-off stall speed of 62 knots. The power-on stall speed would have been something less than this. Once the aircraft had encountered the onset of the stall condition the pilot was left with no alternative other than to lower the nose and accept a forced landing. Although the pilot had a valid reason for not completing a magneto function check prior to takeoff, the contamination of the spark plugs would have shown up during this check had one been carried out.