Fish meal is the most widely used protein source in commercial eel foods and information on the nutritive value of more economical protein sources is needed in order to improve cost-effectiveness of diets. This investigation was conducted to determine apparent digestibility coefficients of available plant and animal meals for juvenile Australian short-finned eel (Anguilla australis australis, Richardson). The suitability of a modified Guelph-type settlement faecal collector in eel digestibility studies was assessed. Animal by-product (spray-dried meat solubles; blood meal; high fat poultry meal) and plant protein (soya bean; canola; maize gluten; lupin; field pea) meals were mixed with a reference diet and marker (0·3: 0·69: 0·01) and the resultant test diets given to the juvenile eels (3·15 (s.e. 0·42) g) at 0·05 live weight per day. The reference diet and all the test diets were well accepted by the fish. Apparent digestibility coefficients for crude protein (ADCCP) for maize gluten meal (MGM), lupin meal (LM) and blood meal (BM) were found to be 0·97, 0·96 and 0·96 respectively and they were significantly (P < 0·001) higher than that for the other ingredients. However, apparent digestibilities for dry matter (ADCDM) and energy (ADCkJ) were significantly (P < 0·0001) higher for animal by-products than for plant proteins except for maize gluten meal. This was explained by the higher content of nitrogen free extractives in the former plant proteins. There was a strong positive correlation between ADCDM and ADCkJ for all ingredients (P < 0·01). Weaker positive correlations were also found between ADCCP and ADCDM and between ADCCP and ADCkJ. Similar results obtained for warm water species using similar faecal collection techniques and over limited eel digestibility data support the suitability of the modified Guelph-type settlement collector system in digestibility studies with juvenile eels.