In this paper, we evaluate query suggestion for Boolean queries in a news monitoring system. Users of this system receive news articles that match their running query on a daily basis. Because the news for a topic continuously changes, the queries need regular updating. We first investigated the users’ working process through interviews and then evaluated multiple query suggestion methods based on pseudo-relevance feedback. The best performing method generates at least one relevant term among 5 suggestions for 25% of the searches. We found that expert users of news retrieval software are critical in their selection of query terms. Nevertheless, they judged the demo application as clear and potentially useful in their work.