Abstract: A medical record processor and method include a voice processor for receiving unprompted and unstructured free dictation, for processing the dictation, and for generating voice data comprising a plurality of terms therefrom. A key term comparator identifies a key medical term in the generated voice data by comparing each term in the generated voice data with a plurality of key medical terms. The identified key medical terms are separated from non-key medical terms in the generated voice data, and a non-key medical term is eliminated from the generated voice data. The identified key medical term is matched with a patient condition, and additional data are extracted from a knowledge database on the patient condition. A draft medical record is generated that includes the identified key medical term and the additional data.