Effects of environmental enrichment at different stages of life on stress physiology of pigs were investigated in a trial with 63 groups, each of four siblings. In each of the three growing phases (suckling 0 to 4 weeks of age, nursery 5 to 9 weeks, fattening 10 to 24 weeks) pens either were (=E) or were not (=0) enriched. Accordingly, the treatments were (i) 000, (ii) E00, (iii) EE0, (iv) 00E, (v) 0EE and (vi) EEE. The enrichment material, renewed twice daily to leave a thin layer, consisted of wood shavings and chopped straw. Salivary cortisol was sampled hourly from 0700 to 1900 h at the age of 9 and 21 weeks. The presence of a circadian secretion rhythm was evaluated by an intra-assay coefficient of variation-based method. An adrenocorticotropic hormone test was performed at 21 weeks. Treatment effects on the odds of a physiological cortisol rhythm were assessed by logistic regression, and effects on cortisol concentrations with a repeated measures GLM. Substrate-enrichment from 0 to 9 weeks of age increased the odds of a rhythm as compared to barren housing (odds ratio (OR) = 30.0, P < 0.01). A flat cortisol secretion pattern may indicate chronic stress and/or delayed maturation of the rhythm. Barren as compared to enriched rearing (0 to 4 weeks of age) seemed to cause a blunted secretion rhythm at 21 weeks of age. Although behavioural and tail lesion observations provided support to the assumption that a blunted rhythm indicates chronic stress, the biological significance of these cortisol results needs confirmation in future studies.