The objectives of the work were (a) to compare the efficacy of two routes for antibiotic administration in the treatment of mastitis in ewes and (b) to assess the potential importance of the timing of the initiation of the therapeutic regime on the outcome of the treatment. The ewes were allocated at random into three equal groups; intramammary inoculation with a Staphylococcus simulans isolate was performed, and clinical mastitis developed. The ewes in groups T1 (n = 6) and T2 (n = 6) were treated by the intramammary administration of ampicillin and dicloxacillin (two administrations with a 12-h interval). The ewes in group T3 (n = 6) were treated by the intramuscular injection of ampicillin and dicloxacillin (0.75 mL per 10 kg bodyweight, three injections with a 24-h interval). In the ewes in groups T1 and T3, treatment started immediately when the clinical signs of mastitis were first detected during the periodic examination of the ewes; in the ewes in group T2, treatment started 24 h after the clinical signs of mastitis were first detected. The animals were monitored clinically; mammary secretion samples were collected for bacteriological and cytological examinations. The median duration of the clinical signs was 4.75, 7.13, and 4.75 d for T1, T2, and T3; significant differences in clinical severity between the groups were seen until the 7th day post-treatment. The median duration of bacterial recovery was 3.25, 8.00, and 8.00 d for T1, T2, and T3; significant differences in the frequency of bacterial recovery between the groups were seen until (64.1%, 94.9%, and 96.2% of the samples) and after (2.9%, 16.7%, and 11.8%) the 7th day post-treatment. The median period required for the complete cure (clinical, bacteriological, and cytological) was shorter in the T1 than in the T2 and T3 ewe groups: 20.0, 32.0, and 24.5 d, respectively. The findings cover a gap in the available literature regarding the treatment of clinical mastitis in ewes. Early treatment resulted in the improved cure of the infection. The comparison of the intramammary and injectable routes for antibiotic administration indicated some benefit for the former, primarily in the post-treatment somatic cell counts.