In this contribution I should like to summarise what for me are currently the main issues of concern about the use of the First-Read-Yesterday (FRY) method in general readership studies. When we talk today about this method, we think mainly about two national readership studies which have adopted FRY, namely the Dutch SummoScanner and the Danish Media Index. The Dutch experiences go back to 1982. Their developments and refinements are documented for international audiences in the books of proceedings of the Readership Symposia of Montreal 1983 Salzburg 1985 and Barcelona 1988. The Danish Media Index has switched to the method in January 1989. What I know about it I have learned at the annual EMRO (European Media Research Organisation) meetings, which I was privileged to attend. At these EMRO meetings I also heard about recent experiments concerning FRY undertaken by CESP in France, and by AG.MA in Germany. I am grateful to EMRO for letting me use the information obtained. Talking about the general use of FRY also implies that CATI (Computer Assisted Telephone Interviewing) is used, as is the case in Holland and in Denmark. The FRY method has therefore two question areas which need addressing if one is asked to consider its possible use for general readership studies: (a) are there any biases arising out of using the questioning technique, known as FRY, and (b) are there any biases arising out of using the telephone? In the following, I will deal with three main concerns. One is the FRY question wording, the second is the problem of small samples generated by FRY for many individual titles, and the third relates to potential biases arising out of the use of the telephone. Before I come to these concerns, I shall briefly outline some chief dissatisfactions with the "recent reading" model and consider whether they are indeed so strong that one might wish to consider FRY as a possible alternative.