In cavity ring-down technique, cavity maladjustment has an essential effect on the measurement of intracavity loss. Several adjustment criterions have been adopted to achieve the optimal cavity state. However, experimental study shows that these criterions may correspond to different cavity states, which means that there is discrepancy between different criterions. In view of this problem, a model of intracavity propagation of Gaussian beam is established based on the angular spectrum propagation theory. This model is tested by numerical simulation and experimental research together. In the simulation, the true value of intracavity loss can be known beforehand. The two-dimensional angular scanning is carried out for certain cavity mirror. The two-dimensional distributions of the measure value of intracavity loss and the transmission light intensity are obtained simultaneously. These distributions are both nonlinear and multi-extremum, which will doubtlessly increase the difficulty in realizing the cavity adjustment. By comparing the distributions , we do find the discrepancy between the largest transmission light intensity and the least measured intracavity loss. Meanwhile both of these two states may be not corresponding to the true value in fact. After statistical studies, the relative error of the least measured intracavity loss is (–37.01±11.79) ppm, whereas the relative error of the largest transmission intensity is (–2.70±0.89) ppm. The criterion of the largest transmission intensity shows better stability and repeatability. This model is further tested in a folded cavity ring-down setup. The similar scanning procedure is carried out. A major problem in the experiment is that the true value of intracavity loss cannot be known. So only the repeatability precision of the measured intracavity loss can be analyzed. The statistical results of the largest light intensity and the least measured intracavity loss are ±29.32 ppm and ±70.71 ppm, respectively. The criterion of the largest transmission intensity has better repeatability, which is basically consistent with the simulation result. In this way the rationality of this model can be verified to some degree. In this paper, the criterion of the largest transmission intensity is recommended in the cavity ring-down technique. Furthermore, this model can be a reference for the research of intracavity optical field response, intracavity optical field transmission, unstable resonator alignment, etc.