The effect of feeding, starvation and fibre ingestion on alkaline phosphatase (ALP) activity (E.C. 3.1.3.1) was studied in Wistar rat serum. Using identical assay conditions for total ALP activity determination and for electrophoretic ALP isoenzymes/fractions activity calculation, alpha- and beta-naphthyl phosphates and p-nitrophenyl phosphate were used as substrates and 2-amino-2-methyl-1-propanol/HCI was used as buffer, respectively. Total activity with beta-naphthyl phosphate was significantly higher than with alpha-naphthyl phosphate and p-nitrophenyl phosphate; with alpha-naphthyl phosphate it was significantly higher than with p-nitrophenyl phosphate. With all substrates, fed animals had significantly higher total activity than starving ones. Electrophoresis allowed the separation of two fractions. The second fraction activity was significantly higher in the fed group than in the starving ones, irrespective of the substrate used. Starving animals with fibre showed higher values of this fraction than starving animals without fibre, the difference reaching statistical significance with alpha-naphthyl phosphate. The first fraction predominated in both starved groups and the second in the fed group. The second fraction was identified as intestinal ALP. We conclude that the mechanical stimulation of the digestive tract appears to influence the passage of intestinal ALP to serum. The experimental conditions used enable quantification of electrophoretic fractions based on total activity. Activity depends on the substrate used.