The adaptive modulation hypothesis posits that the expression of digestive proteins should be modulated in response to intake of their respective substrates. A corollary of this hypothesis suggests that dietary flexibility and digestive plasticity should be correlated. We examined these two hypotheses in two granivorous Chilean birds (Zonotrichia capensis and Diuca diuca) that differ in dietary breadth. D. diuca is a strict granivore, whereas Z. capensis also eats insects. In field-caught birds, the activity of the intestinal dipeptidase aminopeptidase-N was positively correlated with intake of insects in Z. capensis but not in D. diuca. This is the first field documentation of modulation of intestinal enzymes by diet in birds. Intestinal maltase and sucrase activities were not correlated with seed (vs. insect) intake in either species. In the laboratory, captive birds of both species exhibited similar modulation of membrane-bound intestinal hydrolases when fed on synthetic diets of contrasting carbohydrate and protein composition. Maltase, sucrase, and aminopeptidase-N activities were significantly higher in birds fed on the carbohydrate-free than those on the carbohydrate-containing diet. Activities of the three enzymes were positively correlated. Therefore, this increase probably resulted from nonspecific increases of all enzymes resulting from intake of the carbohydrate-free diet. Principal components analysis separating the effect of diet on specific and on nonspecific modulation revealed that diet had a strong effect on nonspecific activity of intestinal enzymes in both Z. capensis and D. diuca. Diet also significantly affected aminopeptidase-N activities when the effect of diet on nonspecific modulation was removed. Birds fed on the carbohydrate-free, high-protein diet had significantly higher specific aminopeptidase-N activities than those fed on the carbohydrate-containing diet. Our results cast doubts on the notion that dietary flexibility and the plasticity of the gut's enzymes are necessarily correlated and on the general validity of the adaptive modulation hypothesis.