The use of tractors for agricultural works showed an important role to mechanized agricultural sector. A repairable
mechanical system (as agricultural tractor) is subject to deterioration or repeated failure. In this study, the regression model
was used to predict the failure rate of MF399 tractor. The machine failure pattern was carefully studied and key factors
affecting the failure rate were identified in five regions of Khouzestan province. The tractors grouped in four sub groups
according to annual use hours and maintenance policies (corrective maintenance and preventive maintenance). Results
showed that the majority of recorded failures were observed in the electrical system (20.3%) for each machine sub-group.
Maximum and minimum annual repair and maintenance costs were related to other (tire, ring, ball bearing and operator seat) and
brake system respectively. According to the results of the research, different annual use hours and maintenance policies
affected failure rate. The tractors included in 300-1,000PM subgroup were mostly in the useful life period while the machines
were towards the wearing out period of machine life for 300-1,000CM subgroup. The tractors maintained under 1,200-2,000PM
subgroup showed an obvious indication of wear-out period entrance, while the tractors were mostly in the wear-out period for
the 1,200-2,000CM sub group. The tractors included in 300-1,000 annual use hours were commonly in a randomized
breakdown period during their useful life but these tractors tend to enter the wear out period in the 1,200-2,000 annual use hours.
So, annual use hours were more effective on failure rate in comparison with maintenance policies.