Frequency of occurrence is an important attribute of lexical units, and one that is widely used in psychological research and theorization. Although printed frequency norms have long been available for Spanish, and subtitle-based norms have more recently been published, oral frequency norms have not been systematically compiled for a representative set of words. In this study, a corpus of over three million units, representing present-day use of the language in Spain, was used to derive a frequency count of spoken words. The corpus consisted of 913 separate documents that contained transcriptions of oral recordings obtained in a wide variety of situations, mostly radio and television programs. The resulting database, containing absolute and relative frequency values for 67,979 orally produced words, is presented. Validity analyses showed significant correlations of oral frequency with other frequency measures and suggest that oral frequency can predict some types of lexical processing with the same or higher levels of precision, when contrasted with text- or subtitle-based frequencies. In conclusion, we discuss ways in which these oral frequency norms can be put to use. The norms can be downloaded from www.springerlink.com.