The effect of different sealing materials \[i.e., polyvinyl chloride (PVC) transparent film, and Parafilm (PARA)\] for Petri dishes was investigated on shoot regeneration from quince (Cydonia oblonga L.) ‘BA 29’ leaf explants. Leaves were excised from proliferating shoot cultures, transversally scored, and placed with the abaxial side down in 60-mm Petri dishes containing 10 ml of Murashige and Skoog modified medium, with 5.4 μM α-naphthaleneacetic acid, 4.5 μM thidiazuron, 200 mg l⁻¹ cefotaxime, and 0.25% (w/v) Phytagel (IM medium) for shoot bud induction, and cultured in darkness at 22 ± 2 °C for 28 d. Then the explants were transferred to standard conditions (16-h photoperiod at 30 μmol m⁻² s⁻¹ photosynthetically active radiation) on a medium similar to IM, except for lack of NAA, and with 0.65% (w/v) agar instead of Phytagel, for an additional 15–28 d. The sealing combinations PARA–PARA, PARA–PVC, PVC–PARA, and PVC–PVC (in the induction–expression phases) were compared during regeneration and for their carry-over effect on shoot development after transfer of explants to an elongation medium (0.9 μM 6-benzyladenine). Carbon dioxide accumulated at 27.2 mmol mol⁻¹ at the end of induction, and gradually decreased from 35.4 mmol mol⁻¹ on day 9 to 22.5 mmol mol⁻¹ on day 28 of the expression phase in PARA-sealed Petri dishes, being always much higher than after sealing with PVC (1–2 mmol mol⁻¹). Ethylene concentration was 0.1 and 0.04 μmol mol⁻¹ in the first part of the induction and expression phase, respectively, in PARA-sealed Petri dishes, and slightly decreased with duration of exposure to light during expression; while it was absent in most PVC-sealed dishes. The PARA–PARA and PVC–PVC (induction–expression) combinations gave, respectively, the worst and best results of regeneration and successive shoot development.