We examined the governing factors of fatigue limit in annealed and austempered ductile iron specimens machined micro hole(dia.<0.4㎜) in rotary bending fatigue test. Also, the quantitative relationship between fatigue limit and maximum defect size in specimens was investigated. Artificial defect(micro-pit type, dia.<0.4㎜) on specimen surface did not bring about an obvious reduction of fatigue limit in austempered ductile iton(ADI) as compared with annealed ductile iron. According to the investigation of {{{{ SQRT { area}_{ c} }} }} which is the critical defect size to crack initiation at artificial defect, {{{{ SQRT { area}_{ c} }} }} o f ADI was larger than that of annealed ductile iron. This shows that the situation of crack initiation at artificial defect in ADI is more difficult in comparison with annealed ductile iron. Maximum defect size is one of the important parameters to predict fatigue limit. And, the quantitative relationship, between the fatigue limit{{{{ { sigma }_{ w} }} }}a n d t he ma x imum defect size{{{{SQRT { area}{} _{ max} }} }}c a n b e e x p r e s s ed to {{{{ { sigma }_{w }{}^{n } · SQRT { area}{} _{ max}= {C }_{2 }. }} }} w h e r e , {{{{n, { C}_{2 } }} }} a r e c o n s t a n t , Moreover, it is possible to explain the difference in fatigue limit between ,austempered and annealed ductile iron by introducing the parameter {{{{ delta (={N }_{sg } /{ N}_{total } ) }} }} i n a plain specimen