Sixteen RCTs were included in the review: 11 (n = 924; 176,332 catheter days) assessed antibiotics; and five (n = 661; 63,345 catheter-days) assessed other antimicrobials. Follow-up ranged from six to 16 months. Seven trials had adequate allocation concealment and allocation sequence generation. Six were double-blind. None of the trials assessed all-cause bloodstream infection. Antibiotic lock solutions significantly reduced catheter-related bloodstream infections compared with heparin (RR per patient 0.44, 95% CI: 0.38, 0.50; RR per catheter-day 0.37, 95% CI: 0.30, 0.47; 11 RCTs). Heterogeneity was significant in both analyses. Subgroup analyses of larger trials with adequate randomisation revealed smaller, but still significantly significant reductions in infections without significant heterogeneity (RR 0.60, 95% CI: 0.54, 0.67; eight RCTs). Catheter removal rates were also significantly reduced by antibiotic lock solutions compared with heparin (RR per patient 0.35, 95% CI: 0.23, 0.55; RR per catheter-day 0.34, 95% CI: 0.21, 0.55; five RCTs) without significant heterogeneity. Non-antibiotic lock solutions also significantly reduced catheter-related infections and catheter removal, again with significant heterogeneity for catheter-related infections. Analysis of funnel plots revealed a significant small-studies effect. Results for other analyses and outcomes were reported.