Summary. Studies in experimental animal models and in patients receiving low molecular weight heparin (LMWH) to prevent thromboembolic events after surgery have not demonstrated a clear relationship between anti-Xa and anti-IIa activities in plasma and either bleeding or prevention of thrombosis. The relationship between these clinical outcomes and ex vivo anti-Xa and anti-IIa activities, activated partial thromboplastin time (APTT) and D-dimers were evaluated in 440 patients undergoing total hip replacement and given prophylaxis once daily with a LMWH (tinzaparin or enoxaparin) in a multicentre doubleblind randomized study. 221 patients received 4500 anti-Xa IU of tinzaparin; 219 patients received 40 mg (4000 anti-Xa IU) of enoxaparin. Both regimens were administered subcutaneously once daily. Blood samples for anti-IIa, antiXa, D-dimers levels and APTT were taken at baseline, on day 1, day 5 and on the day of discharge (days 8-14) and clinical assessments were performed daily until day 14. All patients had bilateral venography between days 8 and 14.All coagulation tests were performed in central laboratories. A significant correlation was observed between anti-IIa activity and anti-Xa activity and the dose of each LMWH injected. The anti-Xa activity was significantly higher with enoxaparin and the anti-IIa activity was significantly higher with tinzaparin. No clear relationship between these two activities and the clinical outcomes was observed. This was also true with regards to APTT. Before and after surgery, Ddimers were significantly higher in patients with deep vein thrombosis (DVT) than in those without DVT but had no predictive value. Interestingly, a significant post-operative increase of D-dimers persisted in both groups of patients during the whole observation period, possibly suggesting that a longer duration of prophylactic treatment may be appropriate.