Even in conversation with neutral semantics, listeners are capable of perceiving a speaker’s positive or pejorative attitude on the content of an utterance through voice and manner of speaking. The present study focuses on the analysis of such evaluative statements and investigates by which acoustic characteristics positive and negative evaluations are distinguished. Considerable differences regarding voice pitch, pitch variation, intonation contour, duration of prosodic units and the manner of stress realization are demonstrated. Positive evaluations are spoken with higher pitch and more pitch variation, evoking the impression of a varied, lilting speaking style. Pejorative prosody, by contrast, is marked by a rather monotonous speaking style, accompanied by a deeper voice, only little changes in pitch and a significantly slower speech rate. Accentuation in pejorative statements is mainly realized by syllable prolongation, while in positive evaluative statements accentuation is indicated by significant pitch ranges.