This study analysed two non-invasive oocyte selection methods in relation to in vitro embryo development capacity and expression of apoptosis-related genes. Selection was based on morphological quality of oocytes or follicle diameter. Oocytes were classified as grade I (GI ≥3 layers compact cumulus cells and homogeneous cytoplasm; grade II (GII ≤3 layers compact cells and homogeneous cytoplasm;, and grade III (GIII ≥3 layers, but cells with slight expansion and slightly granulated cytoplasm). Blastocyst development was lower for GII (28.5%) than for GIII (47.7%, p < 0.05), and GI was similar to both (36.9%, p > 0.05). Relative expression of Bcl-2 gene was lower in the GI (1.0, p < 0.05) than in the GII (1.8) and GIII (2.2), which were not different (p > 0.05). There was no difference (p > 0.05) between GI (1.0), GII (0.92) and GIII (0.93) regarding the Bax transcript. However, the Bax and Bcl-2 transcript ratios in GII (Bax; 0.92 and Bcl-2; 1.8) and GIII (Bax; 0.93 and Bcl-2; 2.2) were different (p < 0.05). Regarding oocytes from follicles of different sizes, cleavage and blastocyst rates for 1-3 mm (82.5; 23.7%) were lower (p < 0.05) than for 6-9 mm (95.6; 41.1%), but similar (p > 0.05) to 3-6 mm (93.7; 35.4%), which were not different (p > 0.05). Regarding Bax and Bcl-2 expression, the oocytes were similar (p > 0.05) for 1-3 mm (Bax; 1.0 and Bcl-2; 1.0), 3-6 mm (Bax; 1.0 and Bcl-2; 0.93) and 6-9 mm (Bax; 0.92 and Bcl-2; 0.91). In conclusion, oocyte selection based on morphological appearance does not guarantee the success of embryonic development. Additionally, the absence of apoptosis is not necessarily a benefit for the development of oocytes. Bovine COCs with initial signs of atresia may be used for the in vitro production of embryos, and COCs taken from follicles >3 mm in diameter are better suited to in vitro embryo development.