BACKGROUND: Evidence from randomized controlled trials has shown that delayed cord-clamping is beneficial to infant iron status. The role of maternal anaemia in this relationship, however, has not been established. OBJECTIVE: To determine the effect of maternal anaemia at delivery on the association between timing of umbilical cord-clamping and infant anaemia at 4 and 8 months of age. METHODS: A cohort of pregnant women admitted to the labour room of Hospital Iquitos (Iquitos, Peru) and their newborns were recruited into the study during two time periods (18 May to 3 June and 6-20 July 2009). Between the two recruitment periods, the hospital's policy changed from early to delayed umbilical cord-clamping. Maternal haemoglobin levels were measured before delivery, and the time between delivery and cord-clamping was recorded at delivery for the entire cohort. Mother-infant pairs were followed-up at 4 (n = 207) and 8 months (n = 184) post partum. Infant haemoglobin levels were measured at follow-up visits. Data were analysed using logistic regression models. RESULTS: The prevalence of maternal anaemia (Hb <11.0 g/dl) at delivery was 22%. Infant haemoglobin levels at 4 and 8 months of age were 10.4 g/dl and 10.3 g/dl, respectively. Infant haemoglobin levels did not differ significantly between infants born to anaemic mothers and those born to non-anaemic mothers at either 4 or 8 months of age. However, the association between the timing of cord-clamping and infant anaemia was modified by the mother's anaemia status. Significant benefits of delayed cord-clamping in preventing anaemia were found in infants born to anaemic mothers at both 4 months (aOR = 0.59, 95% CI 0.36-0.99) and 8 months (aOR = 0.38, 95% CI 0.19-0.76) of age. CONCLUSION: The study contributes additional evidence in support of delayed cord-clamping. This intervention is likely to have most public health impact in areas with a high prevalence of anaemia during pregnancy.