Frequent breakdowns of power tiller, particularly in the peak labor constraints, have been a cause of decreasing machine productivity and its efficient utilization. To keep power tiller always ready to work, it may be necessary to identify the failure characteristics of some major parts of power tiller, and to have those parts replaced accordingly before they show the symptom of mulfunction. This study was conducted to identify the failure characteristics of some major parts of power tiller and to estimate their lives (MTBF) using a Weibull distribution function. The result showed that about half of breakdowns currently being occurred may be reduced through timely replacement of those part having wearout failure characteristics.