Single dose, subconjunctival injections of procaine penicillin using either skin or conjunctival routes of administration were evaluated in order to compare the duration of antibacterial concentrations at the site ofMoraxella bovisinfection. Samples of conjunctival sac fluid (CF) were collected using blunted capillary tubes and were periodically assayed for penicillin by an agar‐well‐diffusion technique. Linear regression lines were constructed for concentration of penicillin‐time profiles. Similarity of the regression slopes for each treatment enabled the construction of three common lines by the method of co‐variance analysis in order to represent each treatment. Treatments were thereafter compared by the slopes of the common regression lines. A subconjunctival injection of procaine penicillin at a dose of 6 ×105iu in 2 ml, administered either through the skin or through the conjunctiva, produced a peak penicillin concentration in the CF of approximately 8 iu/ml for either route and a duration of therapeutic concentration (DTC) of 67.6± 4.9 and 40± 2.7 h respectively. A dose of 3± 105iu in 1 ml, administered through the conjunctiva, produced a similar peak but the DTC was reduced to 35± 4.2 h. The results of this study support field practices, as adequate penicillin concentrations were maintained by both techniques