Narrative:

The first officer was climbing the aircraft with the autoplt roll and pitch channels engaged. He leveled off at FL330, our assigned cruise altitude, and switched on the altitude hold mode. 1 or 2 seconds later the stabilizer trim activated, which by itself is not unusual during autoplt use when airspeed varies. A few seconds after that, the aircraft pitched up, quickly, yet smoothly, and altitude increased to approximately 33250'. At this point, the first officer disconnected the autoplt, but the excessive amount of nose up trim cause the aircraft to balloon up to 33500'. He countered with heavy forward pressure on the control column and nose down trim. Immediately after the climb was stopped, I reported our problem to ZKC while the first officer descended back to FL330. Even though this entire altitude deviation event took place in about 15 seconds, it lasted long enough for the automatic quality assurance program of the center radar to indicate a pilot deviation. There was traffic at FL350 that had passed over us a few seconds before our problem occurred. This traffic had been previously pointed out to us by the center controller and we had it in sight at all times. There was no conflict. This problem was caused by a malfunctioning autoplt, and I don't see how it could have been avoided. We reacted promptly and decisively to the situation and reported it before the controller noticed it on his radar scope. What comes to mind after reflection back upon the whole episode is that whenever you FLIP on an important switch in an airplane, be ready for anything!

Related reports:

Google
 

Original NASA ASRS Text

Title: AUTOPLT FAILURE CAUSED ACFT TO ZOOM GAINING 500'.

Narrative: THE F/O WAS CLBING THE ACFT WITH THE AUTOPLT ROLL AND PITCH CHANNELS ENGAGED. HE LEVELED OFF AT FL330, OUR ASSIGNED CRUISE ALT, AND SWITCHED ON THE ALT HOLD MODE. 1 OR 2 SECS LATER THE STABILIZER TRIM ACTIVATED, WHICH BY ITSELF IS NOT UNUSUAL DURING AUTOPLT USE WHEN AIRSPD VARIES. A FEW SECS AFTER THAT, THE ACFT PITCHED UP, QUICKLY, YET SMOOTHLY, AND ALT INCREASED TO APPROX 33250'. AT THIS POINT, THE F/O DISCONNECTED THE AUTOPLT, BUT THE EXCESSIVE AMOUNT OF NOSE UP TRIM CAUSE THE ACFT TO BALLOON UP TO 33500'. HE COUNTERED WITH HEAVY FORWARD PRESSURE ON THE CONTROL COLUMN AND NOSE DOWN TRIM. IMMEDIATELY AFTER THE CLB WAS STOPPED, I RPTED OUR PROB TO ZKC WHILE THE F/O DSNDED BACK TO FL330. EVEN THOUGH THIS ENTIRE ALT DEVIATION EVENT TOOK PLACE IN ABOUT 15 SECS, IT LASTED LONG ENOUGH FOR THE AUTOMATIC QUALITY ASSURANCE PROGRAM OF THE CENTER RADAR TO INDICATE A PLTDEV. THERE WAS TFC AT FL350 THAT HAD PASSED OVER US A FEW SECS BEFORE OUR PROB OCCURRED. THIS TFC HAD BEEN PREVIOUSLY POINTED OUT TO US BY THE CENTER CTLR AND WE HAD IT IN SIGHT AT ALL TIMES. THERE WAS NO CONFLICT. THIS PROB WAS CAUSED BY A MALFUNCTIONING AUTOPLT, AND I DON'T SEE HOW IT COULD HAVE BEEN AVOIDED. WE REACTED PROMPTLY AND DECISIVELY TO THE SITUATION AND RPTED IT BEFORE THE CTLR NOTICED IT ON HIS RADAR SCOPE. WHAT COMES TO MIND AFTER REFLECTION BACK UPON THE WHOLE EPISODE IS THAT WHENEVER YOU FLIP ON AN IMPORTANT SWITCH IN AN AIRPLANE, BE READY FOR ANYTHING!

Data retrieved from NASA's ASRS site as of July 2007 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.