1. A distinction must be made between the fat embolism syndrome, a clinical entity, and fat embolism demonstrated pathologically, which may be found after death following fracture with no prior evidence of the syndrome.

2. One hundred cases of the syndrome encountered over a period of four years have been studied in detail and the diagnostic criteria have been defined. These include one major feature, four minor features and fat macroglobulaemia.

3. Sixteen of the patients died–eight from severe pulmonary insufficiency of the syndrome, eight from other traumatic causes.

4. The prevention of shock is the best measure for prevention of the syndrome. The role of proteases in the production of shock and the place of protease inhibition in treatment of the syndrome are briefly discussed.

5. For the established case the aim of treatment is to ensure an adequate pressure of arterial oxygen.

1. A diagnostic procedure is described which is specific for fat embolism. It allows identification of the pathological fat globules On a stained slide.

2. The amount of pathological fat can be estimated by comparing the serum triglyceride levels before and after filtration.

3. The method is simple, reproducible and inexpensive.