The result of the study questionnaire indicated that most of the sampled dairy cows were reared under unclean environmental conditions and poor udder preparation. Therefore, it is likely that raw milk might be contaminated from manure, soiled bedding and soil
. In addition, water used for cleaning the milking equipment and washing hands has been associated with potential source of gram-negative staining bacterial contamination in bulk tank milk
 which might hold true in this study. Robinson
 suggested that prior to using detergents, it is essential that the equipment be washed with cold water to remove as much previous milk and dirt as possible followed by washing with warm water to remove fatty deposits. Afterwards, the equipment has to be washed again with warm water and stored in a clean, dry and dust free area. As this research result reveals, this was not practiced by 44.7% of the dairy farm owners.
All of the bacteria identified from the milk samples collected directly from the cows’ udder (CCP-1) were coliform bacteria including, Escherichia coli, Klebsiella pneumoniae, Alcaligenes feacalis and Enterobacter earogenes, with frequency of 47.1%, 26.5%, 14.7%, and 11.8%, respectively. Higher isolation frequencies, especially for Escherichia coli and Klebsiella pneumoniae was observed in the current study as compared to similar studies performed to assess bacteriological quality of raw milk in Ethiopia
[18–20]. This might be due to poor and unhygienic bedding condition in the majority of farms and absence of teat dipping and disinfection practices in the current study. These practices have been known as critical components of mastitis prevention and control program in dairy herds
Bacteria cultivated from farmers bucket at farm level were also dominated by coliform bacteria which accounted 69.2% of the total isolates. Pseudomonas aeruginosa, Citrobacter freundi, and Proteus mirabilis were identified species at this critical control point, in addition to those identified at the previous critical point (CCP-1). Other studies have shown that milk from the bucket at the farm level could be contaminated with gram-negative staining bacteria present on teats, teat canal, udder surface, mastitis udder, residents milking system, and contaminated water used to clean the milking systems
. The apparent dominance of coliform bacteria in this study might be due to the fact that dairy farm owners did not use detergents and disinfectants to wash cows’ udder, which could have significantly reduced the level of coliform and some non-coliform bacteria. Karakök
 reported a similar observation in that use of detergents could reduce contamination of milk by bacteria. A relatively higher number of Pseudomonas aerogenosa and Escherichia coli isolates were detected in the present study than in the reports of Alehegne
, Formm and Boor
, and Godefay and Molla
. It is likely that the psychrotrophic Pseudomonas aerogenosa was originally derived from soil and water. Hence, this organism could gain access into bucket through untreated water used to clean the milking systems or improper udder preparation before milking as well as direct contact with fecal material during milking. The organism has previously generally been mostly isolated from raw milk of bulk thank milk
Bacterial isolates from storage container before transportation at milk collection center were similar to isolates identified at the preceding critical point (CCP-2). This finding was similar to those reported by Bashir and Usman
. Similar to our findings at this particular critical point, Khan et al.
 reported that the dominant gram-negative staining bacteria associated with raw milk samples from bulk tank milk were Escherichia coli followed by Pseudomonas aeruginosa. This might result from poor milking hygiene, unnecessary mixing, transfer of milk from can to can and long milk collection rounds coupled with high ambient temperature.
The detection of additional bacterial species such as Proteus vulgaris and Acinetobacter calcoaceticus which were not isolated in the previous critical control points as well as the increased frequency of isolates like Alcaligenes feacalis and Citrobacter freundi from CCP-4 might be attributed to higher environmental contamination during transportation and/or contamination during waiting along the roadside. Milk in this situation could be exposed to contamination by vehicle dust, since the road is made of gravel. This result is lower in frequency compared with a report by Lues et al.
 from South Africa.
All pasteurized and packed milk samples taken from various supermarkets and restaurants at different shelf life were culture negative for gram-negative staining bacteria. This may be explained by the effectiveness of pasteurization at the processing plant that minimized the chance of postpasteurization contamination. This might also be attributed to a lower level contamination of the milk along the path from cow to the pasteurization plant. As Ashenafi and Beyene
 indicated, microorganisms could survive pasteurization temperature if there is high level of contamination of raw milk. Therefore, keeping such pasteurized milk at room temperature for several hours in retail shops or at home may lead to early spoilage of milk. A high contamination of raw milk implies a higher chance for micro-organisms to survive pasteurization. Psychrotrophic bacteria are important here because many of them can produce extracellular thermostable proteolytic and lipolytic enzymes which can survive pasteurization thus affecting the shelf life and quality of milk and milk products during storage
Overall, 61.1% of the bacterial isolates identified at this particular study area belonged to coliform bacteria. Among which, Escherichia coli (30.2%), Pseudomonas aeruginosa (18.9%) and Klebsiella pneumoniae (17%), were the first, second and third dominant bacterial isolates identified from milk consumed in and around Gondar town, respectively. Most of the bacteria identified here are indicators of poor sanitary and hygienic condition of the farms and transportation system which is similar with previous report by Parekh and Subhash, 2008
. The high presence of Escherichia coli in the milk samples imply that faecal contamination could have occurred and subclinically ill cows might have served as the causes of the microbial contamination. Traditional practices are likely to contribute to the contamination of the milk and proliferation of the micro-organisms. The implication is that there is high risk of acquiring foodborne diseases since 36% of the residents have the habit of consuming raw or unpasteurized milk. It is therefore essential to implement control measures at each critical control points identified in our investigation.