Quarter milk samples from 117 dairy cows with elevated somatic cell counts at the end of lactation were bacteriologically and cytologically analysed 14 days and 7 days before drying off and on the day of dry off. 147 quarters showed subclinical mastitis, 224 with unspecific mastitis, and 93 were healthy. On the day of drying off, the cows were therapied intramammary in all quarters with one of the following long-acting antimicrobial suspensions: No 1, containing 1g Cloxacillin, an oily and dynomilled suspension (Orbenin(R) extra), No 2, containing 0,8g Oxacillin, 200mg Oxacillin-Sodium-Monohydrate, an oily suspension (Stapenor retard(R)), No 3, containing 100 000 I.U. Penethacillin, 300 000 I.U. Benethamin-Penicillin, 100 mg Framycetinsulfat, an oily suspension (Benestermycin(R)), No 4, with 1g Cloxacillin, an oily suspension (Vetriclox(R)). 25 cows were left untreated as a control group. Quarter milk samples were taken from these animals on the day of parturition for bacteriological investigation and on days 7 and 14 after parturition for bacteriological examination and measuring somatic cell count. The antibiotic dry cow therapy, regardless the used preparation, resulted in a significant higher grade of bacteriological curing rate and newinfection-rate compared with untreated cows. The cure rate concerning the somatic cell count in quarters with subclinical and unspecific mastitis before drying off was significantly higher than in untreated ones. Such an effect was not seen in healthy udder quarters. Comparison between the four antibiotic suspensions in relation to the grade of somatic cell count depression and somatic cell count cure rates showed qualitative differences between them: Suspension No1 ranked above No2, No2 above No3, followed by No 4. The results provide strong evidence that the therapy of cows with subclinical and unspecific mastitis or the end of lactation with antimicrobial long-acting suspensions at the time of during off is strongly recommended.