OBJECTIVE: To evaluate day-to-day variability in blood glucose curves (BGCs) generated at home and at the clinic for cats with diabetes mellitus. DESIGN: Prospective study. ANIMALS: 7 cats with diabetes mellitus. Procedures-BGCs generated at home on 2 consecutive days and within 1 week at the clinic were obtained twice. On each occasion, insulin dose, amount of food, and type of food were consistent for all 3 BGCs. Results of curves generated at home were compared with each other and with the corresponding clinic curve. RESULTS: Differences between blood glucose concentration determined after food was withheld (fasting), nadir concentration, time to nadir concentration, maximum concentration, and mean concentration during 12 hours had high coefficients of variation, as did the difference between fasting blood glucose and nadir concentrations and area under the curve of home curves. Differences between home curve variables were not smaller than those between home and clinic curves, indicating large day-to-day variability in both home and clinic curves. Evaluation of the paired home curves led to the same theoretical recommendation for adjustment of insulin dose on 6 of 14 occasions, and evaluation of home and clinic curves resulted in the same recommendation on 14 of 28 occasions. Four of the 6 paired home curves in cats with good glycemic control and 2 of the 8 paired home curves in cats with poor glycemic control led to the same recommendation. CONCLUSIONS AND CLINICAL RELEVANCE: Considerable day-to-day variability was detected in BGCs generated at home. Cats with good glycemic control may have more reproducible curves generated during blood collection at home than cats with poorer control.