Abstract： Objective To compare the differences in clinical efficacy between novolin R and novolog (aspart) in a long period pump therapy. Methods 15 patients with T1DM were treated by novolin R-pump at beginning, and then changed to aspart-pump for more than 3 months (3~18 months). SMBG was done at least 3 times a day. Total daily insulin dose(TDD), pump basal rate (BR), total bolus, frequency of hypoglycemic episodes, HbA1c, incidence(%) of BG fluctuation(≥16.7 mmol/L or ≤3.5 mmol/L)and side-effects were compared. Results The aspart therapy group showed the similar SMBG times, BR doses and HbA1c as compared with novolin R therapy group (all P>0.05). The TDD (0.55±0.06 vs 0.79±0.08 U·kg -1·d -1), frequency of hypoglycemic episodes (0.05±0.01 vs 0.11±0.02 times·d -1), incidence of hypoglycemic (≤3.5 mmol/L) fluctuation (5.3%±0.8% vs 11.4%±1.7%), total bolus doses (0.28±0.03 vs 0.47±0.06 U·kg -1·d -1) significantly reduced (P<0.05) in aspart group than in novolin R group. The pump tubular clog times were more in aspart group (42 times)than in novolin R group (22 times)(P<0.05). Aspart-induced pump tubular clog caused DKA in 2 patients. Discussion This observation showed that while compared with novolin R group, the insulin aspart group reduces significantly the postprandial BG, the frequency of hypoglycemic episodes and insulin dosage, but increases the times of tubular clogs and may cause DKA.