BACKGROUND: Postnatal depression - depression that occurs up to one year after a woman has given birth - is an important and common disorder that can have short- and long-term adverse impacts on the mother, her child and the family as a whole. Recommended treatment for postnatal depression is psychological therapy, and for more severe depression, antidepressants. However, many antidepressants are associated with limited response. Neurosteroid gamma-aminobutyric acid (GABA) receptor positive allosteric modulators have been developed for the treatment of depression, including postnatal depression, and have a different mechanism of action than traditional antidepressants. OBJECTIVES: To assess the benefits and harms of brexanolone, zuranolone and related neurosteroid GABA receptor positive allosteric modulators compared to another active treatment (pharmacological, psychological or psychosocial), placebo or treatment as usual for postnatal depression. SEARCH METHODS: We searched Cochrane Common Mental Disorders' Specialised Register, CENTRAL, MEDLINE, Embase and PsycINFO in January 2024. We also searched two international trials registries and contacted experts in the field to identify the studies that are included in the review. SELECTION CRITERIA: We included randomised controlled trials (RCTs) of women with depression during the first 12 months following childbirth that compared neurosteroid GABA receptor positive allosteric modulators with any other treatment (pharmacological, psychological or psychosocial), placebo or treatment as usual. DATA COLLECTION AND ANALYSIS: We used standard Cochrane methodological procedures. The primary outcomes were depression response, depression remission and adverse events experienced by the mother, nursing baby, or both. The secondary outcomes were depression severity, treatment acceptability, quality of life and parenting- and child-related outcomes. We grouped analyses according to whether the neurosteroid GABA receptor positive allosteric modulator was intravenous or oral. We assessed the certainty of the evidence using GRADE criteria. MAIN RESULTS: We identified six RCTs (674 women); all were placebo-controlled trials. Three studies tested intravenous brexanolone; one, intravenous ganaxolone; and two studies, oral zuranolone. Sample sizes ranged from 21 to 196. All were conducted in the USA. We judged the risks of selection, performance, detection, attrition and reporting biases to mostly be low, although the risk of selection and attrition bias was unclear in two studies. The biopharmaceutical companies which made the drugs sponsored all six included studies. They appear to have had a considerable role in the design and conduct of the studies. Intravenous neurosteroid GABA receptor positive allosteric modulators versus placebo Low-certainty evidence suggests there may be little or no difference in depression response (risk ratio (RR) 1.24, 95% confidence interval (CI) 0.74 to 2.06; I = 78%; 3 studies, 267 women) or remission (RR 1.18, 95% CI 0.59 to 2.38; I = 73%; 3 studies, 267 women) at 30 days (classified in this review as the 'early phase' of treatment: between 0 and 5 weeks from commencement of treatment). There is also probably little or no difference in the number of adverse events affecting the mother (RR 1.02, 95% CI 0.71 to 1.48; I = 46%; 4 studies, 325 women; moderate-certainty evidence). There is low-certainty evidence that there may be little or no difference in depression severity (mean difference (MD) -4.22, 95% CI -8.46 to 0.02; I = 78%; 3 studies, 267 women) in the early phase (at 30 days following commencement of treatment); Hamilton Rating Scale for Depression (HAMD-17) score range 0 to 52. Moderate-certainty evidence suggests lower acceptability than placebo, leading to study dropout (RR 2.77, 95% CI 1.22 to 6.26; I = 0%; 3 studies, 267 women). No studies measured quality of life or parenting- and child-related outcomes. Oral zuranolone versus placebo Moderate-certainty evidence suggests that zuranolone is probably associated with an improvement in depression response (RR 1.26, 95% CI 1.03 to 1.55; I = 13%; 2 studies, 349 women) and remission (RR 1.65, 95% CI 1.22 to 2.22; I = 0%; 2 studies, 349 women) at 45 days from commencement of treatment. Moderate-certainty evidence also suggests that zuranolone probably increases the rate of maternal adverse events (RR 1.24, 95% CI 1.03 to 1.48; I = 0%; 2 studies, 349 women), when all adverse events are considered; the most frequent adverse event was somnolence. Zuranolone is also probably effective in reducing depression severity at day 45 (MD -3.79, 95% CI -5.60 to -1.97; I = 0%; 2 studies, 349 women; moderate-certainty evidence); HAMD-17 score range 0 to 52. Low-certainty evidence suggests little or no difference in terms of treatment acceptability between zuranolone and placebo (RR 0.95, 95% CI 0.50 to 1.81; I = 5%; 2 studies, 349 women). No studies measured quality of life. One study reported the Barkin Index of Maternal Functioning (a validated measure of patient-reported maternal functioning within the first year of childbirth), and found that zuranolone improved maternal functioning at day 45 (MD 7.20, 95% CI 1.42 to 12.98; 153 women), but the certainty of this evidence was low. AUTHORS' CONCLUSIONS: This review provides moderate-certainty evidence that zuranolone probably improves depression response and remission but also increases maternal adverse events compared to placebo. There may be little or no difference in depression response and remission and probably little or no difference in maternal adverse events with intravenous neurosteroid GABA positive allosteric modulators such as brexanolone, compared to placebo. Evidence from this review, alongside current clinical guidelines and reference to evidence from the general adult population, could be used to inform an individualised risk-benefit discussion with women seeking treatment for postnatal depression. However, it is difficult to make recommendations about the use of neurosteroid GABA receptor positive allosteric modulators for the treatment of postnatal depression as no studies have compared them to active treatment.