In two experiments on a total of 95 cows (25 Holstein-Friesian, 64 Schwarzbuntes Milchrind) tests were made to determine the urea space 12 minutes after intravenous administration of a defined amount of urea (HVR12), live weight and dorsal fat thickness, also to determine the body fat content investigated by the estimation of total body water with antipyrine. Further to this followed tests of smiliar nature between visuell respectively palpatorial estimated and measured dorsal fat thickness. The results of the tests were analysed dependent on tests of: number of lactation, state of lactation, breed, day of weighing and condition. The HVR12 is unsuitable for estimation of body fat content of dairy cows. There is no connection between the absolute body fat content and the absolute HVR12. Between the relative body fat content and the relative HVR12 combined with r2 = 0,12 a low significant measure of accuracy is established. The statement of the live weight compared with body fat content is limited. Between the absolute respectively relative body fat content and the live weight combined with r2 = 0, 34 and r2 = 0,16 significant relationships are established. With animals of the 1st and 2nd lactation, also the early and middle lactation, the live weight is unsuitable as a parameter to judge the body fat content. With cows of the 3 rd lacatation and late lactating cows the force of evidence rises increasingly. Compared to HVR12, the live weight is more suitable to estimate the body fat content. The dorsal fat thickness has a direct relationship to the body fat content. Of all three parameters the dorsal fat thickness supplies the best evidence. A significant connection of r2 = 0,62 and r2 = 0,58 can be found between the absolute respectively relative body fat content and the dorsal fat thickness. With increasing number of lactation and length of advanced lactation the force of evidence of dorsal fat thickness will increase in relation to the body fat content. Compared with the live weight, dorsal fat thickness permits cows of the 1st and 2nd lactation as well as early lactating animals a positive judgement of the body condition. The utilization of multiple regresion models promisses - compared with single parameters -no substantial improvement of evidence regarding the body fat content, except the dorsal fat thickness combined with the live weight as opposed to the dorsal fat thickness as a single parameter, a slightly improved accuracy of r2 = 0,62 to r2 = 0,69 is achieved. A certain agreement exists between, throuh extrapolation inveastigated urea space (61,7 %) and the total body water (61,1 %), also the therefrom calculated body fat content. The calculation of the body fat content relating to the extrapolated urea space achieves no improvement with reference to the time and expenditure of work in comparison with the standard method with antipyrine. Between the visuell respectively palpatorial estimated and mesured dorsal fat thickness a highly significant accuracy (r2 = 0,85) is established. Soon after a short training period it is possible to estimate the dorsal fat thickness of dairy cows with sufficiant accuracy. Attention must be payed to negative influence of the estimation accuracy, depending on the condition of the cows. Badly conditioned cows with a dorsal fat thickness <= 10 mm would be overrated, well and very well conditioned cows with a dorsal fat thickness more than 24 mm would be underestimated. To ensure an acceptable estimation of accuracy, the supplementary measure of control is therefore a special significance regarding very lean and very fat cows.
