The exchange flow through tidal inlets generates two-dimensional large coherent vortical structures (2DLCS), that are much broader than the water depth and exist because of the inherent instability of shallow shear flows. These vortical starting jets are critical to the mixing that occurs in the inlet area. Depending on the tidal period T, the width of the inlet W, and the maximum velocity in the inlet UMAX, the mixing will vary from poor exchange to efficient exchange. Here, we present laboratory and numerical experiments that study the formation of the 2DLCS at the mouth of the inlets. Experiments were conducted at large scale, in the shallow flat-bottomed water basin at the Institute of Hydromechanics of the University of Karlsruhe, Germany, which has the capability to generate a sinusoidal flow that simulates a series of tidal cycles. A set of idealized inlets were arranged in the tank, and by varying the tidal period and the maximum velocity, three different types of life-history were obtained (stationary dipole, dipole entrains, and dipole escapes). These types of life-history are defined by the mixing number depending if KW is equal, less or greater than a critical value. The experiments were visualized using color dye tracers. To quantify the shallow water velocity field, the Particle Image Velocimetry (PIV) technique was used. From the PIV data the vorticity field was obtained, and the regions where the vortex formed were identified. Then, a vortex time-evolution analysis was developed using iv physical parameters such as the position on the basin of the vortex, the equivalent diameter, and the maximum vorticity among others. The mixing number accurately predicts the behavior of the vortex for the first cycle on idealized inlets for the subsequent cycles; the structures behave differently than predicted by KW, because the blocking effect of the vortex /formed in the previous cycle. For characteristic times t*  tUWless than about 2, the dipole is attached to the inlet and forms rapidly. For later times, the dipole advects downstream, and slowly dissipates. Numerical experiments are also presented. Comparing the numerical data with the laboratory data, good agreement is reached, but important limitations are identified for the grid resolution and domain size.