The effect of starch and sugars on blood sugar level and renal excretion of sugars and galactitol was investigated. Fifty-nine adult cats were divided into seven dietary groups (carbohydrate content in dry matter): STARCH (29-37% starch, decomposed or raw), SUC (36% sucrose), LAC1 and LAC2 (11 and 28% lactose, respectively), GLUC (40% glucose), GAL (39% galactose) and a carbohydrate-free control diet, FAT. Diet STARCH did not significantly influence postprandial blood glucose level (3.65 +/- 0.68 mmol/l +/- SD, n = 16) compared with diet FAT (3.20 +/- 0.77 mmol/l, n = 14) 1, 3 or 6 h after feeding (weighted means). Diet GLUC led to a steep rise in blood glucose concentration 1 h after feeding (5.08 +/- 0.69 mmol/l, n = 6). Diet SUC induced a mild persistent hyperglycemia without marked postprandial changes (4.52 +/- 0.52 mmol/l, n = 15, weighted mean of 0, 3 and 6 h post-prandially). Diet GAL induced persistent hypoglycemia before and after feeding (2.58 +/- 0.38 mmol/l, n = 13) and considerable postprandial galactosemia (3.26 +/- 1.38 mmol/l, n = 7). In the groups STARCH and FAT, glucose was only detectable in traces in urine, whereas all diets containing sugars led to glucosuria. In group SUC, fructose and sucrose were found in urine and in both lactose groups galactose and lactose were found. Diet GAL led to galactosuria (140 mmol galactose/l). In group LAC1, and especially in group GAL, galactitol was detected in urine. These results point to a rather limited capacity of the cat to metabolize sugars.