Micro-end-milling is a cutting technology that removes redundant material from machined workpieces by small-diameter end mills, and is widely used to manufacture miniature complex parts. During micro-end-milling, the cutting vibration caused by weak tool rigidity and high spindle speed is known as a key factor for decreasing machined quality and accelerating tool failure. This study reports on experiments of micro-end-milling of the straight groove for AISI 1045 steel. The waveform characteristics of acceleration vibration were revealed, the relationship between acceleration and milling parameters were analyzed and two types of relationship models were developed. The results show that, during micro-end-milling of the straight groove, the components of acceleration vibration from largest to smallest are in turn the transverse acceleration α_Y, the feed acceleration α_X and the axial acceleration α_Z. Compared with feed velocity v_f and axial depth of cut a_p, the spindle speed n has the highest influence on cutting vibration. The response surface model of acceleration vibration was shown to have a higher prediction accuracy compared to the power function model and is more suitable for the prediction and control of cutting vibration during micro-end-milling.