This doctoral thesis provides a systematic overview of the possibilities of a more effective sea navigation monitoring in terms of safety of navigation, and environment protection needed due to the increase in the number and size of ships. On the basis of the results gained by the undertaken research, a model of early danger detection has been defined in order to contribute to the effectiveness of the Vessel Traffic Service. The fundamental postulate of the model is based on a separate ship voyage monitoring in the monitored area, with respect to the ship’s interaction with the environment (grounding), as well as to the mutual interaction of all the vessels in the system (collision). With the help of the knowledge from the existing Vessel Traffic Services, expert and scientific literature analysis, surveys performed on VTS operators, and navigation simulator research, the possibilities of drawing up a model of early danger detection have been discovered. For the first time, the model facilitates danger estimation and determines the possibility of an accident beforehand by determining the ship's position, and predicting its position as well as the positions of other ships in the future by the principle of 2D (geographic coordinates) + D (time). By comparing the data of each ship with the entire data base regarding the waterway and sea traffic conditions, potential threats to the safety of navigation and the environment are detected in advance. In the proposed model of early danger detection, the ship's domain is defined as a circle around the ship, and it can be used to define three kinds of domain violations of the monitored ship (there is a violation and it is defined by the time of the entering and exiting the domain, the borderline case of domain violation, and there is no domain violation of the monitored ship). For reason of ship's position prediction in space, a ship's speed prediction model has been developed. A navigation simulator was used to simulate the effects of external hydrometeorological elements (wind velocity, wave height, sea current and the course of the elements) on the speed of the monitored ships, while a neural network was used for the making of the model.