These results are part of the validation study of the Welfare Quality® protocol on growing pig farms. The feasibility and variability of several animal-based welfare parameters were assessed on 30 intensive and 11 extensive Spanish farms. A total of 64496 and 5821 pigs, respectively were evaluated based on 11 criteria related to the 4 welfare principles of good feeding, housing, health and appropriate behaviour. Good feeding, good housing and health parameters were scored at pen or individual level according to a three-point scale ranging from 0 (good welfare) to 2 (poor or unacceptable). Appropriate behaviour was assessed by means of scan sampling of social and exploratory behaviour, qualitative behaviour assessment and human-animal relationship test. Results were expressed as percentage of animals or pens affected and as occurrence of certain behaviours. Differences between farms within the same production system were evaluated by means of logistic models and the significance of the farm effect was fixed at P<0.05. In intensive conditions, the variability of the following parameters allowed the differentiation between farms: moderate (mean±sd=45.5±21.04%) and severe bursitis (4.4±5.66%), manure on less and more than 50% of the body (16.5±13.43%, 3.7±6.37%, respectively), tail biting (0.9±2.02%), scouring (12±18.7%), localised and widespread skin discolouration/inflammation (2.8±4.63%, 3.6±5.97%, respectively), expression of social positive (12.2±5.12%) and negative behaviours (5.4±2.66%), exploratory behaviours (31.3±8.00%), and fearful behaviour (13±19.1%). In extensive conditions, the variability of the following parameters allowed the differentiation between farms: poor body condition (0.7±1.59%), wounds (0.4±0.86%), localised skin discolouration (1.1±1.75%), exploratory behaviour (41.3±23.83%) and fearful behaviour (33±43.2%). Therefore, variability within certain parameters is sufficiently high to allow detection of farms with a particular welfare problem. However, the prevalence and occurrence of animal-based parameters should be carefully considered along with environmental and management conditions in order to avoid misinterpretations.