Objective To evaluate the effect of midazolam on platelet activation in patients with coronary heart disease (CHD) .Methods Brachial venous blood samples 5 ml drawn from 10 healthy adult volunteers and 40 patients with CHD were anticongulated with 3.8 % sodium citrate. Platelet rich plasma (PRP) 3 ml was obtained by centrifngation at 800 rpm for 8 min. Experiment Ⅰ : Ten portions of PRP from healthy adult volunteers served as control group with 1 ml for each (group ⅠC, n = 10). Forty portions of PRP from patients with CHD were randomly divided into 4 groups (n = 10 each): ⅠM0, ⅠM1,ⅠM2 and ⅠM3. In group ⅠC and ⅠM0, midazolam was not added while in group ⅠM1 , ⅠM2 and ⅠM3, midazolam was added with the final concentration at 100, 200 and 400 ng/ml respectively and the PRPs were then incubated for 3 min. The platelet aggregation rote was determined using turbidimetric method. Experiment Ⅱ : After the remaining PRP was eentrifnged at 2000 r/ rain for 5 min, washed platelets (WPs) were obtained. Ten portions of WPs from healthy adult volunteers served as control group with 1 ml for each (group ⅡC, n = 10). Forty portions of WPs from patients with CHD were randomly divided into 4 groups (n = 10 each) : ⅡM0, ⅡM1, ⅡM2 and ⅡM3· In group ⅡC and ⅡM0, midazolam was not added while in group Ⅱ, ⅡM2 and ⅡM3, midazolam was added with the final concentration at 100, 200 and 400 ng/ml respectively and the WPs were then incubated for 3 min. The content of platelet cAMP and were measured by enzyme immunoassay. Results Platelet aggregation rate was significantly higher in group ⅠM0 than in group ⅠC(P<0.01). Compared with groupⅠM0, platelet aggregation rate was significantly decreased in group ⅠM2 and ⅠM3(P<0.01) but there was no significant change in group ⅠM1 (P >0.05). The content of platelet cAMP was significantly decreased while the level of TXB2 was significantly increased in group ⅡM0 as compared with group ⅡC (P < 0.01). Compared with group Ⅱ M0, the content of platelet cAMP was significantly increased (P <0.05 or 0.01) while the level of TXB2 was significantly decreased in group ⅡM2 and ⅡM3(P< 0.01), but there was no significant change in group ⅡM1 (P > 0.05). Conclusion Midazolam 200 and 400 ng/ml can inhibit the platelet activation through increasing the content of platelet cAMP and reducing the production of TXA2 in vitro.