Vitrification could provide a promising tool for the cryopreservation of fish embryos. However, in order to achieve a vitrifiable medium, a high concentration of permeable cryoprotectants must be employed, and the incorporation of high molecular weight compounds should also be considered. The toxicity of these permeable and non-permeable agents has to be assessed, particularly when high concentrations are required. In the present study, permeable and non-permeable cryoprotectant toxicity was determined in turbot embryos at two development stages (F stage-tail bud and G stage-tail bud free). Embryos treated with pronase (2mg/ml, 10 min at 22 degrees C) were incubated in dimethyl sulfoxide (Me2SO), methanol (Meth.) or ethylene glycol (EG) in concentrations ranging from 0.5 to 6M for periods of 10 or 30 min, and in 5, 10, and 15% polyvinylpyrrolidone (PVP), 10, 15, and 20% sucrose or 0.1, 1, and 2% X-1000 for 2 min. The embryos were then washed well and incubated in seawater until hatching. The toxicity of permeable cryoprotectants increased with concentration and exposure time. There were no significant differences between permeable cryoprotectants. However, embryos tolerated higher concentrations of Me2SO than other cryoprotectants. Exposure to permeable cryoprotectants did not affect the hatching rate except at G stage with X-1000 treatment and 20% sucrose. Taking into account the cryoprotectant toxicity and the vitrification ability of cryoprotectant mixtures, three vitrification solutions (V1, V2, and V3), and one protocol for stepwise incorporation were designed. The tested solutions contained 5M Me2SO+2M Meth+1M EG plus 5% PVP, 10% sucrose or 2% X-1000. The hatching rate of embryos that had been exposed to the the vitrification solutions was analyzed and no significant differences were noticed compared with the controls. Our results demonstrate that turbot embryos can be subject to this cryoprotectant protocol without deleterious effect on the hatching rate.