Underwater noise is a form of pollution causing significant concern in terms of environmental status. Shipping is considered the main contributor to the total noise at the global scale, since ship radiated noise can propagate up to tens or hundreds of kilometers. This paper reports on a shipping noise prediction tool based on Automatic Identification System (AIS) data and a normal-mode acoustic propagation code. The tool is set up with realistic environmental information and continuously fed with AIS data, as to periodically produce an instantaneous picture of noise level due to point noise sources present in the area under observation. This produces a time series, that is used for statistical post-processing in order to obtain statistical indicators for a meaningful noise assessment. As a case study, experimental results are reported for a regional observation area in Portuguese waters, off the Southwest Coast of Portugal, crossed by heavy shipping routes. Shipping noise maps are generated for AIS data collected over a 9 month interval. Although the results are yet to be validated by terrain acoustic measurements, this exercise illustrates that a great deal of meaningful information on noise distribution can be obtained by using appropriate acoustic propagation models coupled with live AIS data into a handy noise prediction tool, made available at http://www.shippingnoise.com.