OBJECTIVE: The objective of our study was to examine the management of the interaction between acenocoumarol or phenprocoumon and several antibiotics by anticoagulation clinics and to compare the consequences of this interaction on users of co-trimoxazole with those for users of other antibiotics. METHODS: A follow-up study was conducted at four anticoagulation clinics in The Netherlands. Data on measurements of the International Normalised Ratio (INR), application of a preventive dose reduction (PDR) of the coumarin anticoagulant, fever and time within or outside the therapeutic INR range were collected. RESULTS: The study cohort consisted of 326 subjects. A PDR was given more often to users of co-trimoxazole PDR than to users of other antibiotics. The PDR in co-trimoxazole users resulted in a significantly reduced risk of both moderate overanticoagulation (INR >4.5) and severe overanticoagulation (INR >6.0) compared with no PDR, with odds ratios (ORs) of 0.06 [95% confidence interval (CI): 0.01-0.51] and 0.09 (95% CI: 0.01-0.92), respectively. In co-trimoxazole users without PDR, the risk of overanticoagulation was significantly increased compared with users of other antibiotics. All co-trimoxazole users spent significantly more time under the therapeutic INR range during the first 6 weeks after the course than users of other antibiotics. CONCLUSION: PDR is effective in preventing overanticoagulation in co-trimoxazole users, but results in a significantly prolonged period of underanticoagulation after the course. Avoidance of concomitant use of co-trimoxazole with acenocoumarol or phenprocoumon seems to be a safer approach than management of the interaction between these drugs.