Fifteen female adult capybaras, with initial average body weight (BW) of 32.7 (± 5.8) kg, were kept in individual pens to evaluate effect of supplementation of concentrate feed and its supply time on cecotrophy behavior frequency. The animals were allocated in a completely randomized design, with five animals per treatment, receiving three diets: grass only, grass and grain corn offered in a single meal, and grass and grain corn offered in two daily meals; all cecotrophy acts were recorded. Later, in a second experiment, five capybaras received five levels of urea in their diet: 0, 10, 20, 30, and 40 g urea/100 kg BW, replacing soybean meal as true protein source, in a 5 × 5 Latin square design. The animals were weighed and their blood was collected every 2 weeks. The frequency of cecotrophy (Ps < 0.05) was higher when the capybaras were fed grass only (0.5 ± 0.07 acts/h) than when they received grass and corn in a single meal (0.3 ± 0.05 acts/h) and grass and corn supplied in separate meals (0.1 ± 0.03 acts/h). With increased urea in their diet, the capybaras showed initial signs of chronic intoxication, together with increments in serum urea (r = 0.87, P < 0.05) and a decreasing trend in daily weight gain (r = -0.38, P = 0.06). Therefore, when including concentrate feeds in capybara diet, these must be mixed with roughage in a single meal to avoid high decrease in the frequency of cecotrophy behavior and increase in dry matter intake. The replacement of soybean meal with urea in capybara diet is not recommended.