Information about ice is indispensable to navigation in ice-covered sea areas. For vessels traveling long distances in ice, it is worth planning routes that will reduce fuel consumption and travel time, as well as the risk of ending up in hazardous areas or getting stuck in the ice. In addition to observations on board. there is a multitude of data sources available for seafarers like satellite images, ice model data, weather observations and forecasts. However, it is difficult for a human to take into consideration all the time-varying data parameters when planning a route. In this paper, a prototype system for optimizing routes through the ice field is presented. The system integrates state-of-the-art ice modeling, ship transit modeling, and an enduser system as a route optimization tool for vessels navigating in ice-covered waters. The system has recently been validated on board merchant vessels in the Baltic Sea, and the system's performance has been analyzed statistically using AIS data. Based on the AIS data analysis the mean relative error of the estimated transit time was 0.144 [s/s] with a standard deviation of 0.147 [s/s] for long routes (90-650 km), and 0.018 [s/s] with standard deviation of 0.193 [s/s] for 50 km route segments. (C) 2008 Elsevier B.V. All rights reserved.