Before the year 2000, Mycoplasma synoviae was associated mainly with subclinical respiratory infections in broilers in the Netherlands and was considered to have low clinical and economic impact. The subsequent occurrence of M. synoviae arthritis and amyloid arthropathy, and of eggshell apex abnormalities, has resulted in an increasing demand for M. synoviae-free poultry. Therefore, a cross-sectional seroprevalence study was carried out over a 12-month period during 2005 and 2006. Ten blood samples per farm were generally used because M. synoviae was expected to spread quickly. However, for grandparent and layer breeder stock, 24 to 60 blood samples per house were available from a voluntary M. synoviae monitoring programme. Sera were tested by means of the rapid plate agglutination test (agglutination at dilution > or =1:8 was considered positive). The numbers of farms sampled out of the national total were: broiler grandparent, 53/53; broiler parent rearing, 34/150; broiler parent, 114/300; broiler, 185/800; layer grandparent, 13/13; layer parent, 40/50; layer, 173/1250; and meat turkey, 50/75. The seroprevalence of M. synoviae in commercial poultry was high, especially in commercial layers where it was 73% (95% confidence interval (CI)=67 to 80); in layer and broiler grandparent stock, the seroprevalence was 0% and 10%, respectively, based on sample sizes equal to the population size. In layer and broiler parent farms, the seroprevalence was 25% (95% CI=19 to 31) and 35% (95% CI=28 to 44); in both broiler parent rearing and broiler farms it was 6% (95% CI=0 to 13 and 95% CI=3 to 9); and in meat turkey, the seroprevalence was 16% (95% CI=10 to 22).