Radio audience measurement is normally undertaken by means of quarter- hour-diary research. This type of measurement provides the ordinary 'ratings' that are being used for media planning as well as for programme evaluation. This data provides no information about the quality of radio listening. This paper presents the methodology and results of a research programme carried out in 1983 in the Netherlands. The research had two separate objectives: to determine the reliability (and validity) of the currently used diary measurement and on the other hand to quantify the quality of radio listening particularly activities carried out during listening and the attention level of the listeners.