Rivers have always been the main sources in providing water demand for human kind and the basic element in population development. One of the ways of understanding rivers' behavior is to investigate their flow structure and bedforms. On the other side vegetation exists widely in natural river systems creating roughness on channel walls and flood plains which affects the geometrical shape, hydraulics and flow resistance that follows by sediment traort. Considering the importance of bedforms such as dunes in sediment traort control, turbulence production and resistance to flow, studying the details of interaction of bedforms, vegetated banks and flow structure seems to be necessary. In this study, the influence of straight crested gravel dunes and vegetation cover on channel banks were investigated on flow structure. To reach this purpose seven fixed artificial 2-D straight crested dunes were built along a rectangular flume at the Hydraulics Laboratory of Isfahan University of Technology, Iran. The flume had a length, width and depth of 8m, 0.4m and 0.6m respectively. The graded gravel particles used for building the dunes had a median diameter of d50=10mm and the trapezoidal shape of the dunes were kept constant through the experimental study. The stems of Johnson grass with a median diameter of 2.8mm were used to simulate vegetation cover on the flume walls. With respect to the fact of having fully developed flow located after the fifth dune, experimental measurements were conducted over the fifth and sixth dunes. In total 3 runs were performed over dunes with a wave length, height, angle of repose and flow depth of 0.96m, 0.04m, 28 degrees and 0.28m respectively. In the first case 17 velocity profiles and in the second and the third case 21 velocity profiles were measured. The experiments were all performed with a discharge of 0.024m3/s. The instantaneous three-dimensional velocity components were measured using a down looking Acoustic Doppler Velocimeter (ADV) which was developed by the NORTEK Company. Velocities were recorded for each point with a sampling frequency of 200Hz and the sampling volume of 5.5mm. The sampling duration at each point was 120 seconds. Ultimately more than 45,400,000 velocity data were collected and they were filtered by WinADV software. An important result in this study was that in gravel bedforms with straight crest unlike the sharp crest beforms, in both cases of with and without vegetation cover, no negative velocities were observed. According to the horizontal velocity contour maps, in the case of no vegetation cover, the least velocity value was zero which took place at a small region on the lee side of the dune. Also in the case of with vegetation cover the least velocity value was zero too but its location had moved down stream at the dune's trough. Observing the negative vertical velocity value in both cases of with and without vegetation cover through a dune wavelength was anot