The profile of norethisterone and newly developed derivatives thereof were assessed by in vitro binding and transactivation assays on progesterone (PR) as well as on androgen (AR) receptors and by subcutaneous treatment in in vivo models. The following in vivo models were performed: A McPhail test for progestational activity in immature rabbits, an ovulation inhibition test in cycling rats and a Hershberger test for androgenic activity in immature orchidectomised rats. The compounds tested were: norethisterone (NET), 11-methylene-NET (11-NET), Δ15-NET (15-NET), 18-methyl-NET (18-NET, Levonorgestrel, LNG), 11-methylene-Δ15-NET (11,15-NET), 11-methylene-18-methyl-NET (11,18-NET, 3-keto-desogestrel, Etonogestrel, ETG), (Δ15-18-methyl-NET (15,18-NET, Gestodene, GSD) and 11-methylene-Δ15-18-methyl-NET (11,15,18-NET). Compared to the non-substituted compound NET, the binding to and agonistic activity via PR was increased for all the three mono-substituted compounds, although the stimulatory effect of 15-NET was only twofold. Compounds with 18-methyl in combination with Δ15 (GSD), with 11-methylene (ETG) or with both combined showed clear synergistic effects, leading to equipotent compounds. If the 18-methyl group was lacking as in 11,15-NET, potency was lower than for ETG or GSD, but higher than for 18-NET (LNG). A correlation coefficient of 0.9 was found between binding affinity and agonistic potency. With respect to the AR binding and transactivation activities, the 18-methyl group potentiated androgenic in vitro activity (LNG). The 11-methylene group increased relative binding affinity in NET, but reduced androgenic activity clearly when also other substituents were present (11,15-NET, ETG and 11,15,18-NET). The Δ15 bond alone did not change the binding in NET, but decreased androgen binding, induced by the 18-methyl substituent, in GSD and 11,15,18-NET. Transactivation activity was also diminished in the compounds having a Δ15 bond. In the McPhail test mono-substitution of NET increased the progestagenic in vivo activity three to five times. Bi- and tri-substitution enhanced the activity further. With respect to ovulation inhibition mono-substitution of NET resulted in three to nine times more potent compounds, whereas bi- and tri-substitution increased potency further, except for 11,15-NET, which was as active as 11-NET. The relative progestagenic potencies in the McPhail and ovulation inhibition tests, correlated significantly with those of the relative binding affinity values (correlation coefficient of 0.91 and 0.93, respectively) and relative transactivation activity values (0.88 and 0.81) for the PR. In the Hershberger test, all the compounds increased androgenic activity with respect to growth of ventral prostate weight compared to NET, with the exception of 11,15-NET and 11,15,18-NET. The androgenic activity was negligible for these latter compounds. The androgenicity of both 18-NET (LNG) and 15,18-NET (GSD), on the other hand, was significantly higher than that of 11,18-NET (ETG). The results of this in vivo test are in line with the AR binding and transactivation activity values (correlation coefficients of 0.86 and 0.88). In addition, selectivity indices were calculated by dividing the progestational potencies by androgenic potencies for both in vitro and in vivo assays. ETG and GSD had clearly higher in vitro and in vivo indices than the other compounds with NET and LNG having the lowest indices. Because the androgenicity of 11,15-NET and 11,15,18-NET was very low, no exact selectivity ratios could be calculated for these compounds. From these experiments we may conclude that small structural modifications exert enhancement of progestational activity and a clear reduction in androgenicity leading to very selective progestagenic compounds. The influence of bi-substitution is additive over mono-substitution, whereas tri-substition is not additive. The three substituents (11-methylene, Δ15 bond and 18-methyl) increase the progestational activity over that of NET and combinations of these substituents result in an even further increase in activity. The effect on androgenic activity of the three substituents is more complex: the 18-methyl group similar as in LNG increases the androgenic activity, whereas the 11-methylene group increases RBA in NET, but reduces androgen binding in combination with any other substituent (ETG, 11,15-NET and 11,15,18-NET). The Δ15 bond had no effect in NET, but decreased androgen binding, induced by the 18-methyl substituent, in GSD and 11,15,18-NET. The effects of these eight progestagens in in vitro binding and transactivation studies are very representative for in vivo endometrium proliferation in rabbits, ovulation inhibition in rats and growth of ventral prostate in orchidectomised rats.