In this paper we apply the flow approach to the Dutch labour market using two sets of regional data. From the first set we estimate matching functionsfor unemployed using data on the period 1980-1993. Using the second dataset we compare matching functions forunemployed and employed job seekers using data on the period 1981-1985. We conclude that the efficiency of the Dutch labour market has not worsenedduring the 1980s and that there are no differences in labour market efficiency between regions. Furthermore, weconclude that unemployed and employed job seekers have a different matching technology, while there are no regional differences in the matchingtechnology. Finally, we conclude that the matching function has constant returns to scale.