Background The practice of breastfeeding is related to socio-demograph
ic characteristics: both show marked variation by place within the Uni
ted Kingdom. This study set out to test the commonly held assumption t
hat geographical variations in breastfeeding practices are explained b
y socio-demographic differences. Methods The mothers of 326 six-week-o
ld babies living in Salisbury or Durham Health Districts were intervie
wed to determine attitudes to breeastfeeding practice. Results More mo
thers in Salisbury breastfed than mothers in Durham, both immediately
after birth (Salisbury 79 per cent, Durham 54 per cent) and at six wee
ks (50 per cent and 26 per cent). Higher rates of breastfeeding after
birth were associated with higher level of qualification, encouragemen
t to breastfeed antenatally and the mother having been breastfed herse
lf. Higher rates at six weeks were associated with higher level of qua
lification, higher parity and willingness to breastfeed away from home
. The differences between districts persisted after adjustment for the
se variables, but were diminished when willingness to breastfeed away
from home was taken into account. Conclusion Differences between the t
wo districts in breastfeeding practice cannot be wholly explained by t
raditional socio-demographic characteristics, and may be related to lo
cal culture. Health education is unlikely to change breastfeeding prac
tices unless prevailing cultural attitudes also change.