Objective To review the method used to prepare and evaluate animal models of perimenopausal depression, including animal selection, ovariectomy (OVX), chronic unpredictable mild stress (CUMS) stimulator use, modeling and treatment start time and behavioral testing. Methods We searched the CNKI and PubMed databases for studies of animal models of perimenopausal depression from the establishment of the databases to May 2022. Studies were screened according to defined inclusion and exclusion criteria, the relevant information was extracted and summarized, and charts were drawn. Results A total of 88 articles were identified, including 71 Chinese and 17 English articles. There were 65 rat models and 23 mouse models. The modeling method were OVX+CUMS and OVX+CUMS+solitary culture. Most studies in rats used SD rats (OVX), aged 90 ~ 120 days and weighing > 200 g; most studies in mice used KM mice, weighing about 20 g; 76% of the studies selected seven or eight stimulators, one or two of which were selected randomly each day and were not repeated within 3~ 7 days, with fasting, water deprivation, tail clipping, ice-water swimming, and day night reversal recommended method of simulation. Grouping according to behavioral result reduced the influence of individual differences. Vaginal smears were generally started 3 days after OVX and repeated for 5 days; 89% of the studies chose “treatment and modeling at the same time” or “ treatment after modeling”, and modeling and treatment should be consistent, generally 21 or 28 days. Positive control drugs included fluoxetine or fluoxetine combined with estrogen. The most common behavioral experiments were open field, sugar preference, and forced swimming experiments. Conclusions We reviewed the most widely used preparation method and evaluation indicators for animal models of perimenopausal depression, including details of the modeling processes, to improve the repeatability of preparing animal models of perimenopausal depression.