Real-time PCR has the potential to accurately quantify the mRNA level of selected genes in single cells and individual pre-implantation-stage embryos. The goal of our study was to examine the variations in gene expression within individual embryos of the same stage and between embryos of the same stage but from different sources. In our study, we determined expression level of the 7 most commonly used housekeeping genes in 8-cell-stage mouse embryos produced under different culture conditions. Messenger RNA of 6 embryos each that was derived in vivo, or cultured in vitro from the zygote stage, or derived from oocytes activated parthenogenetically and developed in vitro were extracted individually followed by reverse transcription into cDNA. Optimized real-time PCR was performed for cytoplasmic beta-actin (Actb), glyceraldehyde-3-phosphate dehydrogenase (Gapdh), H2A histone family, member Z (H2afz), hypoxanthine guanine phosphoribosyl transferase 1 (Hprt1), ubiquitin C (Ubc), peptidylprolyl isomerase A (cyclophilin A) (Ppia), and eukaryotic translation elongation factor 1 epsilon 1 (Eef1e1) genes. The results were analyzed, and the percentage standard error of the mean relative expression value was compared for all genes. All 7 genes were presented above the detection limit in all samples. One or two individual embryos showed 2- to 4-fold higher mRNA levels than the average for all genes in the group. The embryos cultured in vitro showed much higher expression levels of H2afz, Ppia, and Eef1e1 genes than those in the in vivo group. The parthenogenetic group was similar to the in vivo group in expression of Actb, H2afz, Hprt, and Eef1e1 genes, but showed significant differences (P < 0.05; Student's t-test) compared to the in vitro group (Table 1). The percent standard error of the mean decreased gradually as the number of samples was increased. The 6 individual embryos in similar groups showed relatively low variability compared to embryos at similar stage but produced in different conditions. Interestingly, the parthenogenetic embryos showed a level of gene expression comparable to that of the in vivo ones, notwithstanding their culture in vitro. In conclusion, morphological observations and similarity in developmental stage alone cannot guarantee the uniformity of embryo samples, and a minimum of 4–6 replicates per treatment is needed. Moreover, we showed that culture condition itself has an effect on housekeeping gene expression, which, if neglected, might result in misinterpretation of data. Table 1.Relative expression values of the different culture groups (mean ±SE; n =6 embryos) This work was supported by EU FP6 (MEXT-CT-2003-509582 and 518240), Wellcome Trust (Grant No. 070246), and Hungarian National Science Fund (OTKA) (Grant No. T046171).