Objectives:To explore the effects of multi-component exercise training on physical function and cognitive functions of the community-dwelling older adults with mild cognitive impairment (MCI).Methods:This study was an experimental study.90older adults with mild cognitive impairment were selected from Wangyuehu community and were divided into control (n=45) or experimental group (n=45) randomly. Participants in the control group were given community conventional health education, and those in the experimental group received multi-component exercise training based on community conventional health education. The participants were assessed with the Chinese version mini-Physical Performance Test (CM-PPT), Mini Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA) before and at the end of3rd,6th months of the intervention. SPSS17.0was used to establish the database, and repeated measures ANOVA was appliedd to explore the effects of multi-component exercise training on physical function and cognitive functions of the elderly with MCI.Results:(1) The baseline data of90older adults with MCI showed that the scores of CM-PPT was11.36±2.69, the scores of tanding static balance was3.51±0.69and the time of physical function domains of CM-PPT were9.91±2.98(s)(chair-stand time),6.54±1.41(s)(6meters timed walk),2.16±1.02(s)(picking up a penny from the floor) respectively. The baseline data of older adults with MCI showed that the scores of MMSE and MoCA were26.57±1.39and21.33±2.01, respectively. The scores of the seven cognitive domains of MoCA were3.14±1.01(visuospatial/executive subscale),2.63±0.53(naming subscale),5.08±0.90(attention subscale),2.26±0.64(language subscale),0.75±0.70(abstract subscale),1.70±1.36(delayed recall subscale) and5.70±0.74(orientation subscale), respectively;(2) Spearman correlation analysis indicated:CM-PTT and MoCA, visuospatial/executive function, delayed recall subscale were related positively(r=0.426,P<0.01; r=0.278, P<0.05; r=0.333, P<0.01); five times chair-stand time and MoCA, language, delayed recall subscale were related negatively(r=-0.411,P<0.01; r=-0.233,P<0.05; r=-0.391,P <0.01);6meters timed walk and visuospatial/executive function, MoCA, delayed recall were related negatively (r=-0.411,P<0.01, r=-0.242,P<0.05; r=-0.295, P<0.01);(3) CM-PPT,MMSE and MoCA scores before and after the intervention were used for the repeated measures analysis of variance.The CM-PPT score with main intervention effect and the CM-PPT scores with the interaction between the intervention and time were both statistically significant (P<0.05).With the exceptions of the MMSE score with main intervention effect, the MoCA score with main intervention effect and the MMSE and MoCA scores with the effect of time, the MMSE and MoCA scores with the interaction between the intervention and time were all statistically significant (P<0.05).Conclusions:(1) There were different levels of damages in many physical function and cognitive domains in community-dwelling older adults with MCI, such as delayed recall, visuospatial/executive function and abstract.(2) Multi-component exercise training could significantly improve physical function and cognitive function of community-dwelling older adults with MCI.