Granulated polyamide (PA) was tested for use as an external marker to estimate faecal DM (FDM) excretion of Zebu cattle (Bos indicus). The study was conducted in Mali, using seven and eighteen animals respectively in four field trials and six indoor experiments. Cattle ate fresh or dry pasture vegetation and half the animals were additionally supplemented with crop byproducts. Gelatine capsules containing 35, 40 or 45 g PA were administered orally at 12 h intervals. Estimates of FDM were based on the average marker concentration in faeces and were correlated with the actual excretion measured by total faecal collection. The pre-measurement period required to establish equilibrium for regular marker dosing was determined at 4 d. Except for diets with a N content of less than 9·26 g/kg organic matter, marker recovery averaged 98·1 (se 0·93) % (n 62), and was not influenced by diet composition and the quantity of feed ingested (P > 0·05). Estimates of FDM based on average PA concentrations in faecal samples were correlated to the actual excretion with r 0·98 (n 62; P ≤ 0·001). Since the PA concentration in individual faecal grab-samples is not correlated with either sample mass or sampling time, accurate estimates of FDM require a grab-sampling schedule that covers the 24 h day. However, estimates of FDM were found to be acceptable if calculations are based on the average PA concentration in the sub-total of samples collected during the day or during night respectively (r 0·95, n 29; P ≤ 0·001 in both cases). It is concluded that the use of PA marker is a simple and inexpensive method resulting in reliable estimates of FDM. Since sophisticated analytical procedures are not required to recover PA in faecal samples, the marker is particularly suitable for application in extensive grazing systems and in studies conducted in less-developed countries.