Vascular occlusion is used to reduce blood loss during liver resection surgery. There is considerable controversy regarding whether vascular occlusion should be used or not during elective liver resections. The method of vascular occlusion employed is also controversial. There is also considerable debate on the role of ischaemic preconditioning before vascular occlusion. To assess the advantages (decreased blood loss and peri-operative morbidity) and disadvantages (liver dysfunction from ischaemia) of vascular occlusion during liver resections. To compare the advantages (in decreasing blood loss or decreasing ischaemia-reperfusion injury) and disadvantages of different types of vascular occlusion versus total, continuous portal triad clamping. We searched The Cochrane Hepato-Biliary Group Controlled Trials Register, the Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library, MEDLINE, EMBASE, and Science Citation Index Expanded until March 2007. We included randomised clinical trials comparing vascular occlusion versus no vascular occlusion during elective liver resections (irrespective of language or publication status). We also included randomised clinical trials comparing the different methods of vascular occlusion and those investigating the role of ischaemic preconditioning in liver resection. We collected the data on the characteristics of the trial, methodological quality of the trials, mortality, morbidity, blood loss, blood transfusion requirements, liver function tests, markers of neutrophil activation, operating time, and hospital stay. We analysed the data with both the fixed-effect and the random-effects models using RevMan Analysis. For each binary outcome we calculated the odds ratio (OR) with 95% confidence intervals (CI) based on intention-to-treat analysis. For continuous outcomes, we calculated the weighted mean difference (WMD) with 95% confidence intervals. We identified a total of 16 randomised trials. Five trials including 331 patients compared vascular occlusion (n = 166) versus no vascular occlusion (n = 165). Six trials including 521 patients compared different methods of vascular occlusion. Three trials including 210 patients compared ischaemic preconditioning before continuous portal triad clamping (n = 105) versus no ischaemic preconditioning (n = 105). Two trials including 127 patients compared ischaemic preconditioning before continuous portal triad clamping (n = 63) versus intermittent portal triad clamping (n = 64).The blood loss was significantly lower in vascular occlusion compared with no vascular occlusion. The liver enzymes were significantly elevated in the vascular occlusion group compared with no vascular occlusion. There was no difference in the mortality, liver failure, or other morbidities. Four of the five trials comparing vascular occlusion and no vascular occlusion used intermittent vascular occlusion. Trials comparing complete inflow and outflow occlusion to the liver, ie, hepatic vascular exclusion and portal triad clamping demonstrate significant detrimental haemodynamic changes in hepatic vascular exclusion compared to portal triad clamping. There was no significant difference in the number of units transfused and the number of patients needing transfusion. There was no difference in mortality, liver failure, or morbidity between total and selective methods of portal triad clamping. All four cases of mortality and liver failure in the comparison between the intermittent and continuous portal triad clamping occurred in the continuous portal triad clamping (statistically not significant). Intermittent portal triad clamping does not increase the total blood loss or operating time compared to continuous portal triad clamping.There was no statistically significant difference in the mortality, liver failure, morbidity, blood loss, or haemodynamic changes between ischaemic preconditioning versus no ischaemic preconditioning before continuous portal triad clamping. Liver enzymes used as markers of liver injury were significantly lower in the early post-operative period in the ischaemic preconditioning group. The intensive therapy unit stay and hospital stay were statistically significantly lower in the ischaemic preconditioning group than in the no ischaemic preconditioning group.There was no statistically significant difference in the mortality, liver failure, morbidity, intensive therapy unit stay, or hospital stay between ischaemic preconditioning before continuous portal triad clamping and intermittent portal triad clamping. The blood loss and transfusion requirements were lower in the ischaemic preconditioning group. Aspartate aminotransferase level was lower in the intermittent portal triad clamping group than the ischaemic preconditioning group on the third post-operative day. There was no difference in the peak aspartate aminotransferase levels or in the aspartate aminotransferase levels on first or sixth post-operative days of aspartate aminotransferase . Intermittent vascular occlusion seems safe in liver resection. However, it does not seem to decrease morbidity. Among the different methods of vascular occlusion, intermittent portal triad clamping has most evidence to support the clinical application. Hepatic vascular exclusion cannot be recommended routinely. Ischaemic preconditioning before continuous portal triad clamping may be of clinical benefit in reducing intensive therapy unit and hospital stay.