1. The effect of the haemorrheological agent pentoxifylline was investigated in a canine model of acute myocardial infarction, induced by occlusion of the left anterior descending coronary for 6 h. Thirty minutes post-occlusion the dogs were randomized to receive either distilled water or pentoxifylline (0.3 mg kg-1 min-1 for 1 h followed by 0.15 mg kg-1 min-1 for 4.5 h) intravenously. 2. At 6 h post-occlusion the in vivo area at risk was determined with monastral blue dye and the area of necrosis was determined with triphenyltetrazolium chloride. The area at risk was 16.5 +/- 1.3% in the control group (n = 10) and 17.2 +/- 1.8% in the pentoxifylline treated group (n = 10; NS). The area of necrosis was 12.3 +/- 1.9% in the control group and 11.9 +/- 2.2% in the pentoxifylline treated group (NS). The area of necrosis expressed as a percentage of the area at risk was 69.3 +/- 7.7% in the control group and 63.6 +/- 7.4% in the pentoxifylline treated group (NS). 3. Pentoxifylline had no significant effects on heart rate, systolic or diastolic blood pressure. Regional myocardial blood flow, measured by the radioactive microsphere technique, was not significantly different between the groups. 4. Thus, pentoxifylline does not reduce infarct size in this model of acute myocardial infarction and does not enhance coronary collateral blood flow.