No objective evidence has been presented to support the beneficial effect of physical training on glycaemic control in Type 1 (insulin-dependent) diabetic patients trained two to three times a week for several months. In the present study we examined the possibility that a daily exercise programme would be more suitable for improving glycaemic control. Thirteen patients completed a 5-month study; 6 were randomized to exercise training (20 min daily bicycle exercise) and 7 served as non-exercising controls. The training resulted in an 8% increase in maximal oxygen uptake (p less than 0.05). No change in glycaemic control occurred during the study period in either group. In addition, serum lipid and lipoprotein levels were followed. Total cholesterol decreased during the study period irrespective of training. No effect was noted on the levels of LDL, VLDL, HDL and HDL2 cholesterol. A significant training effect was obtained in the HDL3 subfraction (-10%, p less than 0.05). Total triglycerides were unchanged, but a decrease in the level of LDL triglycerides was observed with training (-12%, p less than 0.01). It is concluded that, in female Type 1 diabetic patients, daily physical training for several months does not improve glycaemic control and results only in minor changes in serum lipoprotein profiles.