Unexpectedly, we found no overall beneficial effect on mortality in a randomised trial of vitamin A supplementation (VAS) or placebo administered with BCG vaccine at birth in Guinea-Bissau. We conducted an explorative analysis to examine whether subsequent diphtheria-tetanus-pertussis (DTP) vaccinations had modified the effect of VAS at birth. VAS was associated with a weak tendency for decreased mortality as long as BCG was the most recent vaccination, the mortality rate ratio being 0.86 (0.48-1.54); 0.82 (0.32-2.08) in girls and 0.89 (0.43-1.88) in boys. However, after DTP vaccination VAS at birth was associated with increased mortality in girls (2.19 (1.09-4.38)), whereas no difference was seen for boys (0.90 (0.44-1.82)) (p=0.08 for equal effect of VAS in the two sexes if DTP is the last vaccine). The explanation for the lack of beneficial effect in our setting may have been that VAS at birth interacted negatively with subsequent DTP vaccinations in girls.