Introduction: Clinical studies have shown that acupuncture relieves different kinds of pain [1-2]. However, the effectiveness of these methods is unclear. Previously, we reported brain activity with short-term manual acupuncture stimulation [3-6]. The curative effects of the acupuncture last even after treatment. Resting-state functional MRI (rs-fMRI) allows detection of resting state networks (RSNs), which represent areas that display functional connectivity (FC) at “rest.” Therefore, in this study, we used rs-fMRI and FC analysis to examine the RSN activity before and after acupuncture stimulation. Methods: MRI acquisitions: Thirteen healthy subjects (men, 9; women, 4; age rage, 18-33 years) underwent MRI examination with a SIEMENS 3.0-T Trio MRI system having a 32-channel head array coil. All subjects provided written informed consent. rs-fMRI images for each subject were acquired using a gradient echo type-EPI (GRE-EPI) sequence with TR/TE = 2000/30 ms, voxel size = 3.6 x 3.6 x 3.5 mm3, FOV = 224 mm, gap = 0 mm, 36 slices, for 185 volumes and a total acquisition time (AT) of 6.18 min. A high resolution 3D T1-weighted scan was also performed using a magnetization-prepared rapid gradient-echo (MPRAGE) sequence (TR/TE/TI = 1900/2.52/900 ms; flip angle = 9°; 176 sagittal slices; voxel size = 1 x 1 x 1 mm3, FOV = 256 mm) for anatomical reference. Experimental design: A specific acquisition protocol comprising 4 consecutive rs-fMRI scans (labelled PRE, INSERT, POST1, and POST2 as shown Fig.1) was designed to study dynamic changes in RSNs in response to the acupuncture stimulation performed during the interval between with the second and third rs-fMRI acquisition. The acupuncture stimulation consisted of block designs with five 15-s stimulation blocks (on) interspersed with one 30-s and five 45-s rest blocks (off). Stimulation was applied by bidirectional needle rotation through approximately 180°, with even motion at a frequency of 1 Hz. fMRI analysis: The first 13 data points of each dataset were discarded to obtain the stable state. Volume registration, head motion correction, blurring were performed by SPM 8. ICA analyses were carried out using MELODIC (FSL, version 5.0.5)[7]. A non- parametric permutation test (dual regression[8]) was then applied to create and compare group-specific maps for each independent spatial component, using 500 permutations in the contrasts design. This step allowed detecting statistically significant differences in RSNs from the pre (PRE) and post acupuncture groups (INSERT, POST1, and POST2) were tested using a non-parametric 2-sample t-test. Statistical maps uncorrected with a threshold of p≤0.01 for statistic significance. ACUPUNCTURE STIMULATION CHANGES DMN FUNCTIONAL CCNECTIVITY. Results & Discussion: Fig 2 shows the demonstrated increased activity (uncorrected. p < 0.01, red-yellow voxels) in default mode network (DMN). In INSERT-PRE and POST2-PRE, the needle was inserted and resting state, areas of stimulus-induced activation were observed in the posterior cingulate cortex (PCC) and precuneus. Conversely, in POST1-PRE, the needle was inserted and after manipulated state, area of stimulus-induced activation was observed in the medial prefrontal cortex (mPFC). In the result, DMN was changed with or without manual acupuncture. The acupuncture stimulation controls a pain by modifying mechanism of pain transmission. Generally, the lateral nociceptive system involved in pain discriminative is received the projection from outside thalamic nucleus, and the medial nociceptive system involved in pain cognition and affect pain is received the projection from inside thalamic nucleus. Therefore, we suggest that the resting acupuncture impinge on outside thalamic nucleus, and manual acupuncture impinge on inside thalamic nucleus.