This work investigates the German version of the Kalamazoo Communication Skills Assessment Form (KCSAFd) for three assessment methods: students' self-assessment (KCSAFd-self), assessment by standardised patients (KCSAFd-sPat) and video-assessment by trained raters (KCSAFd-video). Videotaped simulated patient consultations of N = 163 medical students from the first (n = 97) and the final clinical years (n = 66) were rated using the KCSAFd. Investigating the psychometric properties of the instrument, we found a two factor-construct with interpersonal and conversational competence. All methods showed good internal consistency and acceptable model fit values. Additionally, we found plausible relationships between the three methods and meaningful differences between the two groups of students.