Both aspects of the two way interaction between cancer and the haemostatic system were investigated. In a prospective study, Fibrinogen, Fibrinopeptide A (FPA), Fragment Bβ 15-42, Fibrin Plate Lysis Assay and Fibrin(ogen) Degradation Products (FDPs) were measured in patients with operable breast cancer (BC) patients with benign breast disease (BBD) and healthy subjects. Preoperatively, FPA and FDPs were highest in BC patients, but were also significantly elevated in BBD patients. Bβ 15-42 was elevated equally in these two groups. Neither pre nor postoperative haemostatic measurements were of any value in predicting early recurrent disease. Elevated FPA values persisted in BC patients 3 & 9 months postoperatively, whilst Bβ 15-42 rose further during this time. An association between oestrogen receptor result and Bβ 15-42 values was found. These findings suggest that much of the activation of haemostasis in cancer patients arises from non-specific causes, and that haemostatic changes do not correlate with prognosis. They suggest that a primary tumour may cause relative suppression of the fibrinolytic response. The inhibition of metastasis by warfarin was studied in an animal model. Warfarin was not cytotoxic for Mtln3 tumour cells, but inhibited metastasis. Deposition of fibrin within tumours was apparently not altered by warfarin treatment, and injection of fibrin or FDPs with tumour cells had no effect on metastasis or growth of the tumour. Pre-injection warfarin treatment of the host inhibited metastasis of Mtln3 cells, whilst pretreatment of tumour cells had no effect. Injection of factors II, VII, IX and X reversed this effect of warfarin, if given within 12 hours of tumour cells. Further studies demonstrated that factors II, IX and X together enhanced metastasis in non-anticoagulated rats. Finally, Arvin defibrination did not abolish this effect, and did not itself affect metastasis. It was concluded that the factor II, IX, X complex can enhance metastasis, and that the antimetastatic effect of warfarin appeared to be due to inhibition of these proteins. Since enhancement was not affected by defibrination, it may occur via mechanisms other than fibrin formation.