Due to the marked cytoplasmic opacity of canine oocytes, the diagnosis of their nuclear status is difficult. The objective of the present study was to evaluate the accuracy of Hoechst staining observed under epifluorescence wide-field microscopy [living oocyte observation (LivOO)] by comparison to a reference technique [DNA staining with ethidium homodimer-2 under confocal microscopy; fixed oocyte observation (FixOO)]. Four Hoechst 33342 concentrations (200 ng, 500 ng, 1 μg, 2 μg/mL) were tested and 1 μg/mL was the lowest one with the lowest proportion of oocytes in which DNA was missed. At this concentration, LivOO procedure did not affect the degeneration rate. On 379 oocytes observed individually with the two techniques successively, diagnosis of meiosis resumption by LivOO was exact in 87.3% of the cases, but the meiosis resumption rate was underestimated (23.5% versus 34.3% with FixOO; p < 0.001). Diagnosis for metaphase II was exact in 80% of the cases, but LivOO detected only 72.7% of the metaphase II oocytes present. Metaphase rates did not differ between LivOO and FixOO. This study contributes to a better interpretation of in vitro maturation results. The developmental potential of metaphase II canine oocytes sorted after Hoechst staining is to be evaluated.