In order to elucidate the developmental basis for smolting, Atlantic salmon, Salmo salar, parr (< 11.5??cm) and smolts (> 12.5??cm) were exposed to natural daylength (LDN) and increased daylength (LD16:8) starting in late February and gill Na+,K+-ATPase activity and circulating hormone levels monitored from January to May. Gill Na+,K+-ATPase activity remained low and constant in both groups of parr. In smolts, gill Na+,K+-ATPase began increasing in late February in both photoperiods, but was significantly higher in the LD16:8 group from March through April. Smolts exposed to LD16:8 had dramatically elevated plasma GH within one week of increased daylength that remained high through April, whereas plasma GH of LDN smolts increased steadily beginning in late February and peaking in late April. Plasma GH levels of parr remained low in spring and did not respond to increased daylength. Plasma insulin-like growth factor I (IGF-I) levels were substantially higher in smolts than parr in January. Plasma IGF-I levels of parr increased steadily from January to May, but there was no influence of increased daylength. In smolts, plasma IGF-I of LD16:8 fish initially decreased in early March then increased in late March and April, whereas plasma IGF-I of LDN smolts increased steadily to peak levels in early April. Plasma cortisol was low in parr throughout spring and did not differ between photoperiod treatments. Plasma cortisol of LD16:8 smolts increased in early March and remained elevated through April, whereas in LDN smolts plasma cortisol did not increase until early April and peaked in late April. Plasma thyroid hormones were generally higher in smolts than in parr, but there was no clear effect of increased daylength in parr or smolts. The greater capacity of the GH/IGF-I and cortisol axes to respond to increased daylength may be a critical factor underlying smolt development. ?? 2007 Elsevier B.V. All rights reserved.