One hundred samples of feeding-bottles were collected from infants and young children comingrnto four clinics in Addis Ababa, from November, 1996 to April, 1997. Information on sex, age,rnnumber of bottles used, educational status of the mother and ingredients of bottle content wererncollected through interview. The bacteriology of these samples was analysed using standardrnmicrobiological techniques.rnThe analyses showed that all the samples had heavy bacterial contamination, the mean count beingrnin the levei of 107 c.f.u / ml. Only 17% of the samples had count l'O6 c. f. u /ml. The two mostrn*commonly encountered bottle contents (cow's milk and cereal blend) were the most heavilyrncontaminated. About 52% of the milk and 93% of cereal blend had count more than 107 c.f.u/ml.rnThe contamination level was found to be affected by ingredients of bottles used and educationalrnstatus of mothers. About 90% of the samples prepared by illiterate mothers had count over 107rnc.f.u/ml. The percentages for elementary and high school educated mothers for the correspondingrnload were 85 and 79, respectively.rnThree hundred and sixty-nine bacterial isolates, grouped into 12 genera were identified from thernbottle content. Thirty-seven per cent of the samples had 3 isolates, whereas 21% and 16% of thernsamples were contaminated with 4 and 2 isolates, respectively. The dominant organisms wererncoliforms (34%) followed by Staphylococci (28 %), Bacillus spp . (19%) Micrococcus spp. (14%)rnand other (3%). Three isolates of Salmonella spp. of the same serogroup were also identified.rnCow's milk harboured 219(59%) of the isolates followed by gruel made of cereal blend(16%).rnA total of 30 factory-produced weaning foods (product A, B and C), and 20 of the two mostrnfrequently encountered home-made bottle-contents (cow's milk and cereal blend) were preparedrnat home by mothers under close supervision. The bacteriological analyses of these foods showedrnthat only 30% of the factory-produced weaning foods had bacterial count over 102 c.f.u/ ml(g). Therncount from the two home-made bottle contents was also low; only 15% of the samples hadrnbacteria 102 c.f. u/ml. These signifies that food handling and the gap between consumption andrnpreparation must be important in this regard. The organisms isolated from the above 50 samplesrnwere Bacillus spp., for they were the only organisms that could tolerate the heat treatment.rnThe growth potential of Salmonella sp. in products 'C' and cereal blend was determined. It wasrnfound out that it reached to a level of 107 c.f.u ./ml(g) in 12 hours and 10s c.f.u/ml(g) in 16 hours.rnIf weaning foods have initial contamination of 103 c.f.u./ml (g) (which by itself is beyond thernthreshold level for infants) it reaches to unacceptably high level within 8 hours (10G c.f.u/mlg)In order to intervene with food borne diseases of infants and children, mothers must be taught ofrnfood safety principles. Increasing the number of bottles used to feed infants to three or more couldrnalso be helpful in reducing food borne bacterial contamination . More important is teaching andrnencouraging mothers to consider the use of fermented food products as alternatives, which arerncost-effective means of reducing bacterial pathogens in weaning foods.