The authors calculated the effect size for post-operative analgesia of three additives, clonidine, neostigmine, and tramadol to bupivacaine, ropivacaine, or levobupivacaine used for single-dose caudal extradural blockade in children. A meta-analysis was performed for three end points of efficacy: the increase of time until administration of analgesic drugs, the proportion of patients requiring analgesic drugs during the initial 24 post-operative hours, and the amounts of post-operative analgesic drugs. A Bayesian inference supporting direct statements about the probability of the magnitude of an effect was used to compare the effects size. Neostigmine increased the duration of analgesia by 9.96 h (95% confidence interval: 7.75 to 12.16), as compared with 3.68 h (2.65 to 4.7) with clonidine and 4.45 (2.84 to 6.07) with tramadol. There is a 95% probability that neostigmine increases the duration of post-operative analgesia by more than 8 h, clonidine by more than 2.8 h, and tramadol by more than 3.25 h, as compared with local anesthetics alone. The odds ratios for the proportion of patients requiring analgesic drugs were 0.22 [0.13 to 0.37] for clonidine and 0.28 [0.10 to 0.75] for neostigmine. With tramadol, there was no statistically significant difference. All three additives reduced the amounts of post-operative analgesic drugs. Neostigmine and tramadol increase the probability for post-operative nausea or vomiting (PONV). Neostigmine provides the longest post-operative analgesia. With clonidine, the duration of analgesia is shorter and sedation is increased, but the probability for PONV could be decreased.