We constructed a model of apoptosis in mouse preimplantation embryos and investigated the effect of the flavonol icariin on embryonic development in vitro in embryos with reduced microRNA-21 (miR-21). The model was generated by microinjecting an miR-21 inhibitor into the cytoplasm of mouse pronuclear embryos, which were cultured in vitro using modified CZB (mCZB) basal medium (model group), or using mCZB medium with 0.6 μg/mL icariin as an experimental group (model-Ica). These were compared with embryos collected in vivo (vivo group) or not microinjected (control group). Developmental rates in vitro of two- and four-cell embryos and blastocysts were observed, and Hoechst 33342 and terminal deoxynucleotidyl transferase dUTP nick end labeling staining were used to count blastocyst cell numbers and apoptotic cell numbers and percentages. The transcriptional levels of miR-21, the apoptotic genes caspase 3 and phosphatase and tensin homolog deleted on chromosome ten (PTEN), and the antiapoptotic gene Bcl-2 were detected by quantitative polymerase chain reaction (qPCR). Western immunoblotting was used to detect the protein levels of caspase 3, PTEN, and Bcl-2. Compared with the model group, icariin treatment significantly improved blastocyst development in vitro (58.43 ± 7.53% vs. 37.85 ± 6.47%; P < 0.01), whereas it was not significantly different to the control group (60.34 ± 9.86%). Icariin treatment significantly increased the blastocyst cell numbers (47.02 ± 4.93 vs. 37.70 ± 5.80; P < 0.01), and reduced the rates of apoptosis (5.51 ± 2.35% vs. 10.11 ± 4.21%; P < 0.01), whereas the blastocyst cell numbers and apoptotic rates revealed no significant differences between the vivo (46.06 ± 6.50, 5.95 ± 2.56%) and control groups (45.77 ± 4.09, 6.18 ± 2.41%). Icariin treatment significantly improved miR-21 expression in all embryo stages, reduced the transcriptional levels of caspase 3 and PTEN, and increased the levels of Bcl-2. The protein expression levels of caspase 3 and PTEN were decreased in blastocysts and the level of Bcl-2 was increased (P < 0.01). These had no significant differences with the vivo and control groups, and the protein levels revealed no significant differences between two- and four-cell embryos. Thus, miR-21 was necessary for preimplantation embryonic development, and embryo quality was closely associated with the apoptosis-related protein expression levels regulated by miR-21. Icariin upregulated miR-21 expression and reduced apoptosis in embryos with reduced miR-21. It also improved embryonic developmental quality in vitro, indicating an important regulatory role for miR-21 in blastocyst development in vitro.