Health care professionals often need to convey good and bad prospects to patients, and these news can be qualified by various uncertainty terms. Based on a sociolinguistic analysis of the way these uncertainty terms are used, we predicted that they would be interpreted differently by patients as a function of whether they qualified good news or bad news.Two studies investigating causal inferences were conducted among a sample of French university students (Study 1, N=50), and among a sample of Italian pregnant women (Study 2, N=532).Participants felt greater confidence in the conclusions they derived when the news were bad, as compared to the conclusions they derived when the news were good.The findings have implications for health care professionals who communicate good and bad prospects to patients, and who need to qualify the certainty of these prospects.Professionals should be aware that when the news are bad, any hedging term such as "possible" can be misunderstood as an attempt to sugar-coat the pill, and that this misinterpretation can lead patient to inferences that are not shared by the professional.