We studied the breeding bird density of wader and gull species for 18 years (1967-1984) at a study plot at Finse (1200 m a.s.l.) in the northwestern part of the 10 000 km2 high mountain plateau Hardangervidda, South Norway (60°36’N, 7°30’E). The study plot was a 1 km2 large sedimentation flat in the bottom of the Finse Valley. Eight species of waders were recorded as territorial, and their combined densities varied from 2 to 19 (mean 10) territories/km2. Four species, the Temminck’s Stint, Calidris temminckii, the Common Sandpiper, Actitis hypoleucos, the Ringed Plover, Charadrius hiaticula and the Dunlin, Calidris alpina, occurred regularly (>half of the study period) and constituted about 87% of the annual community densities in the study plot. One species, C. temminckii, was the dominant species, and held on average 60% of the territories. Four species, the Redshank, Tringa totanus, the Purple Sandpiper, Calidris maritima, the Golden Plover, Pluvialis apricaria, and the Red-necked Phalarope, Phalaropus lobatus, occurred only irregularly, with a low number of territories, and did not breed every year. Only one gull species was recorded as territorial, the Common Gull, Larus canus, and the density varied from 1 to 5 (mean 2.4) territories/km2. The variations in density of the wader community was correlated positively with the mean temperature in June, and negatively with the NAO index for June, the date of snow-melt and the percentage of snow cover in the area, showing that an early snow-melt was important for the establishing of territories. Both the wader community, the combined wader and gull community, each of the species C. temminckii, C. alpina and P. lobatus fluctuated in synchrony with the populations in the passerine community of the area. The wader community, the combined wader and gull community and the species C. temminckii and L. canus also varied with the fluctuations of the small rodent populations in the area. The densities of territories of the different wader species are compared with literature data from other parts of their distributional area.