Background: Medications or mechanical dilators are often used to soften and dilate the cervix prior to surgical evacuation of the uterus for non-viable pregnancy, or miscarriage. The majority of miscarriages occur in the first trimester. The aim of cervical ripening is to reduce the possibility of injury to the uterus and cervix and improve the surgical ease of the procedure. Cervical ripening agents can have adverse effects and it is uncertain as to whether these risks outweigh the benefits of their use. Objectives: To systematically review the benefits and harms of using cervical ripening agents prior to surgical evacuation of non-viable pregnancy prior to 14 weeks' gestation. Search methods: We searched the Cochrane Pregnancy and Childbirth Group's Trials Register (30 April 2015) and reference lists of retrieved papers. Selection criteria: Randomised controlled trials (published in full-text form, or as abstracts only), which assessed the use of pharmacological or mechanical agents to ripen the cervix in women undergoing dilation and curettage or vacuum aspiration for non-viable pregnancy at less than 14 weeks' gestation were eligible for inclusion. Cluster-randomised controlled trials and trials using a cross-over design were not eligible for inclusion. Unpublished randomised controlled trials and quasi-randomised trials would have been eligible for inclusion but none were identified. Data collection and analysis: Two review authors independently assessed the studies for inclusion, assessed risk of bias and carried out data extraction. Data were checked for accuracy. Main results: We included nine trials with 469 women. A diverse set of medications and regimens were studied in these trials, making the comparisons available for meta-analysis limited. The comparisons draw data from six trials with 383 participants. All trials were relatively small and had several aspects of unclear risk of bias with few of this review's outcomes reported. Due to this, no data from three trials were able to be used despite them meeting inclusion criteria. We carried out four comparisons: isosorbide mononitrate or dinitrate compared with misoprostol; misoprostol compared with placebo; chemical dilation (use of medications) compared with mechanical dilation; and any cervical preparation compared with placebo. None of the included studies reported data on the review's primary outcome: cervical or uterine injury (perforation, laceration, creation of a false passage). No clear difference was shown between isosorbide compounds and misoprostol for the outcome need for manual cervical dilation (average risk ratio (RR) 0.76, 95% confidence interval (CI) 0.10 to 5.64; three trials, 150 women; Tau2 = 2.11; I2 = 69%), however the data were heterogenous. In terms of adverse effects, misoprostol was associated with more vomiting (RR 0.11, 95% CI 0.01 to 0.85; two trials, 120 women), however there were no clear differences between isosorbide compounds and misoprostol in relation to other reported adverse effects (headache, nausea or hypotension). The dosing regimens differed in terms of dose, number of administrations and route of administration in the different trials. Mechanical (Dilapan-S hygroscopic) dilators performed similarly to chemical dilators in a single trial (65 women) that measured difficulty in cervical dilation, excessive bleeding and adverse effects. Misoprostol was shown to be more effective than placebo for cervical ripening (reduced need for manual cervical dilation) (RR 0.14, 95% CI 0.08 to 0.26; one trial, 120 women), and surgical time was reduced when misoprostol was used (mean difference (MD) -3.15, 95% CI -3.59 to -2.70; one trial, 120 women). However, compared to placebo, misoprostol, was associated with more abdominal pain (RR 29.00, 95% CI 1.77 to 475.35; one trial, 120 women), although no clear differences in the risk of other adverse effects (nausea, vomiting, headache or fever) were observed between groups. There was no clear differences between chemical dilation and mechanical dilators for the outcomes: difficulty in cervical dilation, excessive bleeding or adverse effects. Compared with placebo, any cervical preparation reduced the need for manual cervical dilatation (average RR 0.25, 95% CI 0.07 to 0.89; two trials, 168 women; Tau2 = 0.67; I2 = 81%), and reduced surgical time (MD -2.55, 95% CI -3.67 to -1.43, two trials, 168 women; Tau2 = 0.63; I2 = 96%). None of the included trials reported on the review's other secondary outcomes, including: injury to bladder or bowel, miscarriage/preterm birth in a subsequent pregnancy, analgesia use after administration of ripening agent but before surgery, or analgesia use after surgery. Authors' conclusions: This review found no evidence to evaluate cervical ripening prior to first trimester surgical evacuation for miscarriage for reducing the rate of cervical or uterine injury, however, this may be because these outcomes are very rare. Cervical preparation was shown to reduce the need for manual cervical dilatation compared with placebo. Misoprostol and isosorbide mononitrate and dinitrate were similarly effective in ripening the cervix, however there was more vomiting with misoprostol. Mechanical (Dilapan-S hygroscopic) dilators performed similarly to chemical dilators. The nine studies included in this review were small and the methodological quality of the trials was mixed, and for the most part, not well-described; thus any conclusions drawn from the data included in this review must be treated with caution. Consequently, large, high-quality trials are required to determine whether the benefits of this treatment outweigh the risks. Further research should be powered to assess the rate of cervical and uterine injury between interventions. Future research should also guide clinicians in deciding whether the benefits of reduced manual cervical dilatation outweigh the risks of adverse effects associated with these agents (nausea, vomiting, headache, fever, diarrhoea and pain). Women's satisfaction and outcomes of future pregnancies should also be assessed.