By means of statistical calculations on plant records of a large number of 50 m sections in 5 habitat types (and based on selection of indicative plant species), it was concluded that the survey route within 1 x 1 km-squares to achieve a reasonable result varies from 1-7 km for different habitat types. In squares with little variation in habitats a route length of 2.5 km seems sufficient, whereas in more varied squares a route of at least 5 km is required for a reasonably complete survey.