Dynamic soaring is the rationale behind the prolonged flights of a seabird Albatross. It involves utilization of energy from the wind shear present near the earth surface. Small unmanned aerial vehicles (UAVs) can be kept loitering without any external power input by dynamic soaring. In this work, dynamic soaring is used to power UAVs for the surveillance application. A set of 6-DoF point mass equations governing the aircraft motion is used in the optimal control problem formulation. Appropriate constraints considering the material properties of a UAV, and the loiter pattern of dynamic soaring, are imposed on state variables and control parameters. Trajectories are optimized by using GPOPS-II, MATLAB based optimal control software. The problem is optimized for the efficacy of area under surveillance. Variation in the surveillance area is analyzed with the change in the view angle of camera, wind strength, and nature of wind shear profile. Surveillance by dynamic soaring becomes effective with the increase of wind strength and also with the change of wind shear profile towards the logarithmic variation. The minimum requirement of wind strength to perform dynamic soaring has been identified by considering various wind shear profiles. Finally it is concluded that small UAVs (comparable with the size of Albatross) can be constantly kept on surveying using wind energy as the sole power source, as long as free stream wind velocity is greater than the minimum requirement for dynamic soaring. © 2016 The Authors.