STUDY OBJECTIVENeuraxial hydromorphone provides postoperative pain relief. However, the magnitude of this effect and the optimal dose remain unknown. The objective of this study is to clarify these uncertainties.DESIGNSystematic review and meta-analysis with trial sequential analysis.SETTINGPostoperative recovery area and ward, up to 24 h.PATIENTSAny patient undergoing any type of surgery or being in labor.INTERVENTIONSNeuraxial hydromorphone versus control.MEASUREMENTSOur primary outcome was rest pain score (analogue scale, 0-10) at 24 h according to route of administration (epidural versus spinal) and type of surgery (orthopedic versus other). Secondary outcomes included rest pain score at 0-4 and 8-12 h; rates of postoperative nausea and vomiting, and pruritus at 24 h.MAIN RESULTSSix trials, including 436 patients, were identified. Rest pain score at 24 postoperative hours was significantly reduced in the hydromorphone group, with a mean difference (95 %CI) of -0.4 (-0.8 to -0.1), I2 = 74 %, p = 0.01. Neuraxial hydromorphone did not increase postoperative nausea and vomiting (risk ratio [95 %CI]: 1.2 [0.8-1.8], I2 = 27 %, p = 0.47), but increases pruritus (risk ratio [95 %CI]: 3.1 [1.6-5.9], I2 = 0 %, p = 0.0005). The quality of evidence was very low for our primary and secondary outcomes. In conclusion, there is very low level of evidence that neuraxial hydromorphone provides effective analgesia after surgery or labor, at the expense of an increased rate of pruritus. The improvement in pain scores appears to be clinically insignificant. With only six trials published over a period of 30 years, we were unable to perform a meta-regression.CONCLUSIONSIf neuraxial hydromorphone is to be used regularly, trials focusing on the optimal dose and side-effects should be performed before widely administering this medication into the neuraxial space. More trials focusing on the optimal dose and side-effects should be performed before widely administering this medication into the neuraxial space.