Speakers with Autism Spectrum Disorder (ASD) have been claimed, and are generally assumed, to produce atypical intonation. Previous research, however, is very limited and findings are contradictory, with claims ranging from “robotic” to “singsongy” intonation styles in ASD. We employ a novel method to assess intonation styles in a large corpus of semi-spontaneous speech by German adults with ASD and matched controls. We show that the ASD group has a more melodic and less “robotic” intonation style as a whole, but that this effect is much stronger for males than for females and is subject to individual variation. This finding is related to the observation that females are more likely than males to successfully use compensation strategies in order to hide their specific autistic traits and thereby adapt to majority norms in what has been described as “social camouflage”.