A simple three-step synthesis of 5-(3-hydroxyphenyl)-2-methyl-2-azabicyclo[3.3.1]nonan-4-ol (3a) was achieved using an osmium tetroxide mediated oxidation of the known intermediate 6. A pyrrolidine-ring variant of 3a (3-(7-(hydroxymethyl)-6-methyl-6-azabicyclo[3.2.1]octan-1-yl)phenol (5)) was isolated when other routes were used. The epimeric hydroxy analogue 4a was synthesized by simple inversion of the stereochemistry at C-4. Both N-methyl (3a and 4a) and N-phenethyl (3b and 4b) derivatives were synthesized. The compounds were examined for their opioid receptor affinity and the N-phenethyl analogue 3b was found to have relatively weak affinity for the μ-opioid receptor (K(i) = 74 nM). However, the N-phenethyl analogue of the C-4 epimer, 4b, had about 15 fold higher affinity than 3b and was selective for the μ-opioid receptor (K(i) = 4.6 nM). Compound 4b was a moderately potent μ-opioid antagonist (K(e) = 12 nM), as determined by [(35)S]GTP-γ-S assays. Compounds 3b and 4b were energy minimized at the level of B3LYP/6-31G*, and then overlaid onto the 5-phenylmorphan, the (1R,5R,9S)-(-)-enantiomer of 2b (Fig. 1) with the α or β-OH group at the C-9 position. The spatial orientation of the hydroxyl moiety in 3b, 4b, 2a, and 2b is proposed to be the structural requirement for high μ-opioid receptor binding affinity and their agonist or antagonist activity. The modest change in spatial position of the hydroxyl moiety, and not the N-substituent, induced the change from potent agonist to an antagonist of moderate potency.