Modern pollen-vegetation relationships have been explored in recent years through analysis of either pollen trap (= absolute results) or moss polster contents (= relative results). In the past few attempts were made to compare data from these two sources. It was the aim of the present study to achieve this comparison through more exact sampling of moss polsters. A pilot study was carried at Morton Lochs, Fife, Scotland where pollen traps have been sampled continuously for ten years. Moss samples were collected from cushions adjacent to two of the pollen traps during May 1985, 1986 and 1987. Moss polsters were sampled usingan equivalent surface cross-section area to that of the traps. In comparing the pollen results from the two sources some similarities can be identified, but in general there are more inconsistencies. These may be traced to difficulties in the precision of moss sampling, variable rates of moss growth and local deposition factors. It is considered that further study, in which these difficulties are minimised, would be justified. Ultimately it is hoped that precise sampling of mosses will produce sufficiently reliable modern pollen rain results to enable pollen-vegetation studies to move away from relative (percentage) values without requiring long-term artificial pollen trapping techniques.