In this paper, we predicted abiotic stress responsive microRNAs in wine grape genome by homolog search based on computational approach. We successfully predicted 18 grape microRMAs from total 12 query-microRNAs families which chosed from Arabidopsis, Oryza and Populus plants. The MiPred software prediction results, the stem-loop structure feature analysis and pol II promoter prediction results showed that these 18 putative microRNAs are real plant microRNAs. Several known cold, drought stress responding transcript factor binding motifs were identified in these 18 promoter regions. Word spyanalysis displayed 2 remarkable consensus motifs in these promoters with Arabidopsis microRNAs. Four groups of targets were predicted in this work, including CCAAT-binding transcription factor, ring finger transcription factor, sugar metabolism and transport proteins, methyl-transferase and some unknown grape proteins. 3 putative targets of vv-mirRNA169j, 477a and vv-miR393a have reverse transcriptase and leucine-rich repeats (LRRs) ribonuclease inhibitor activity.