After elaborating a suitable scheme to assess consciousness/unconsciousness of turbot, the effects on fish welfare of commercial and experimental slaughtering techniques were evaluated. Strong adverse reactions and slow loss of clinical reflexes were observed when fish were slaughtered by bleeding without prior anaesthesia. The efficiency of electricity (150 V for 2 s followed by 25 V for 5 min) as a stunning and killing technique was dependent on the current frequency and mode of administration, ie whether the electrical discharge was applied to the head only or throughout the whole body. Finally, percussion of the head with a mechanical device resulted in immediate and permanent loss of all responses and reflexes in all fish tested. Three methods were selected as a result of the preliminary tests (bleeding in ice slurry, whole-body electrical treatment and percussion), and their influence on flesh quality was studied over a 9 day period. Fish killed by percussion were characterised by a higher pH and higher water content in the very early stage of post mortem storage, but also by a much delayed rigor mortis. In contrast, fish killed by electricity entered most rapidly into rigor mortis; their flesh was significantly softer (p < 0.05) throughout the entire storage time and was also redder and darker, as indicated by higher a* values (p < 0.001) and lower L* values (p < 0.01) respectively. It is suggested that automated percussion is a suitably humane method for the slaughter of farmed turbot.