A quasi-experimental study with separate samples for pre and post-test was conducted by means of a structured questionnaire at Maternal and Child and Charoenkrungpracharug Hospital, Bangkok which were both baby friendly hospitals from March 7-April 21, 1993, to identify some of the factors related to mothers perception and practice in breast-feeding. The data was collected the first time before coaching mothers at Postpartum ward and the second time after coaching mothers at Well Child Clinic. An experimental group that consisted of 51 mothers was selected from one hundred and seven mothers involved in breast-feeding from Maternal and Child Hospital. The other selected sample was from Charoenkrungpracharug Hospital and used as a control group which consisted of 56 mothers. The sample groups were interviewed regarding socio-demographic factors like age, education, occupation, residence, marital status, family income, family size

communication media available and visits to ANC

other factors on the perception of mothers like perceived susceptibility of infants to catch infection when not breast-fed, perceived severity of non breast-fed infants illness, perceived benefits/barriers in breast-feeding

factors regarding practice of mothers like nursing position, latch on position of holding the baby, time to give breast-feeding,suckling methods of the baby, burping the baby, cleaning nipple, breast engorgement, etc. Statistical analyses like frequency distribution, t-test group, paired t-test and simple correlation were used to identify the relationship between the two groups. The results of the study revealed through t-test that the association between mothers perceived susceptibility, perceived severity, perceived benefits/ barriers and mothers practice were not significantly different between the two groups

paired t-test between mothers perceived susceptibility, perceived severity, perceived benefits/ barriers and mothers practice were significantly different in both groups

and simple correlation that mothers perceived benefits/ barriers were correlated with mothers practice in breast-feeding. There were significant differences in practices observed between the two groups at time 2 (l-month of delivery). In addition, the experimental group had performed significantly better than the control group in breast-feeding ( (p = .000). - . 000 ) . The conclusion drawn was that coaching had directly improved the practices of postpartum women (p = .000) in the experimental group with little or no effect on the perceived susceptibility (p = .869), perceived severity (p = .809) and perceived benefits/ barriers (p - .010) of those women in breast-feeding.