Synchronised ferret winter pelage development induced by melatonin was utilised as an in vivo model for studying gene expression associated with the hair growth cycle. The effects of melatonin and steroids on hair follicle growth initiation in flank skin samples were monitored by histology and immunocytochemistry. Melatonin implantation initiated hair growth within 10 days. Exogenous steroids delivered by slow-release subcutaneous implants had mainly inhibitory effects on spontaneous and melatonin-initiated hair growth, either when given simultaneously or prior to melatonin administration. 17β-oestradiol totally suppressed hair follicle growth in all circumstances, while the inhibitory effects of dexamethasone, testosterone and deoxycorticosterone were weaker and attenuated by melatonin. Progesterone given simultaneously with melatonin inhibited hair growth in males, but did not significantly affect female hair growth initiated by melatonin. Gene expression underlying ferret hair growth was studied using an optimised differential display technique using two sets of complementary skin samples covering all growth initiation stages. The differential display band patterns for approximately 8% of all skin transcripts suggested that more than a thousand genes are required for hair follicle growth. Of the differentially expressed genes detected in a set of flank skin samples covering all proanagen stages, up to a half are likely to be specifically associated with hair growth. The 150 expressed sequence tags (ESTs) discovered represented 112 unique sequences. Forty-two of these were aligned by sequence homology to known genes, while the remainder are likely to represent novel genes. Differential expression through the follicle growth cycle or after the treatments was confirmed for 21 out of 23 genes whose mRNAs were detectable by Northern blot hybridisation. Overall, expression alteration for most of these ESTs occurred prior to morphological changes in the skin associated with hair follicle reactivation. All the ESTs whose expression sites were detected by in situ hybridisation were localised to hair follicles. Some were detected only in certain stages of the hair follicle cycle, others in all growth stages, but at different levels. Classification of the identified ESTs suggested that apoptosis as well as fatty acid synthesis and cholesterol metabolism were closely associated with hair growth initiation, while cell interaction and motility were also particular prominent processes in hair follicle regrowth. Additional DNA sequence information was obtained for ferret hair acidic keratin 8 (fHa8) and ferret cyclin D-interacting myb-like protein 1 (fDMP1). Both of these genes were expressed only in cortical cells in the keratogenous zone from mid-proanagen. fHa8 was expressed only in one side of the cortex, a feature has not previously been reported for other type I intermediate hair keratins. The restricted expression of fDMP1 and its smaller size suggested that it was an alternatively spliced form of the human and mouse homologues with a specialised function in the regulation of cortical cell proliferation and differentiation. Further studies on the expression regulation and function of the genes represented by the ESTs identified in this study promise to considerably advance our understanding of hair follicle growth control.