The corneal permeability to and metabolism of four phenyl substituted prostaglandin analogues have been studied in vitro. Porcine corneas were mounted in incubation chambers dividing each chamber into an epithelial and endothelial side compartment. The analogues were added to incubation medium on the epithelial side. The permeability coefficients of 17-phenyl-18,19,20-trinor-PGF2 alpha-1-isopropyl ester (PhDH100A), 15-keto-17-phenyl-18,19,20-trinor-PGF2 alpha-1-isopropyl ester (PhXA12), 13,14-dihydro-15-hydroxy (R, S)-17-phenyl-18,19,20-trinor-PGF2 alpha-1-isopropyl ester (PhXA34) and 13,14-dihydro-17-phenyl-18,19,20-trinor-PGF2 alpha-1-isopropyl ester (PhXA41) were determined to be in the range of 5.1-11.0 x 10(-6) cm x s-1. All analogues in the endothelial compartment had been hydrolysed to corresponding acids but any other metabolism of PhDH100A, PhXA34 and PhXA41 after 4 h of incubation was minimal. In contrast, PhXA12 free acid was extensively metabolised to the 13,14-dihydro metabolite. To investigate whether the porcine ocular tissues contain 15-hyroxyprostaglandin dehydrogenase (15-PGDH) activity, prostaglandin F2 alpha (PGF2 alpha) and PhDH100A were used as substrates. PGF2 alpha and the phenyl-substituted analogues were also tested for their capacity as substrate to 15-PGDH in general. The 15-PGDH activity was low in all ocular tissues. The capacity of various ocular tissues or purified 15-PGDH to metabolise PhDH100A was lower than with PGF2 alpha as substrate. PhXA34 and PhXA41 were found not to be metabolised by 15-PGDH. Thus, the phenyl substituted PG esters penetrated the cornea and in the process were hydrolysed to their corresponding acids. No appreciable further metabolism occurred except for PhXA12 which was reduced by delta 13-reductase.