Glycemic control in patients with chronic kidney disease (CKD) is particularly hard to achieve because of a slower insulin degradation by the kidney. It might modify the long-acting insulin analogue pharmacokinetics, increasing its time-action and the risk of hypoglycemia. However, because this insulin has no peak action, it might be a more tolerable approach to patients with CKD. This hypothesis remains to be tested, because no study has thus far examined the efficacy and safety profile of long-acting basal analogues in patients with significant loss of renal function. The purpose of this study was to compare the glycemic response to treatment with glargine U100 or neutral protamine Hagedorn (NPH) insulin in patients with type 2 diabetes mellitus (T2DM) and CKD stages 3 and 4.Thirty-four patients were randomly assigned to glargine U100 or NPH insulin after a 2-way crossover open-label design. The primary end point was the difference in glycosylated hemoglobin (HbA) and in the number of hypoglycemic events between weeks 1 and 24, whereas secondary end points included changes in glycemic patterns, weight and body mass index, and total daily dose of insulin. HbA was determined by ion-exchange HPLC, and hypoglycemia was defined as glucose concentration of 54 mg/dL (3.0 mmol/L) detected by self-monitoring of plasma glucose or continuous glucose monitoring.After 24 weeks, mean HbA decreased on glargine U100 treatment (-0.91%; P < 0.001), but this benefit was not observed for NPH (0.23%; P = 0.93). Moreover, incidence of nocturnal hypoglycemia was 3 times lower with glargine than with NPH insulin (P = 0.047).Our results found that insulin glargine U100 could be effective, once it improved glycemic control, reducing HbA with fewer nocturnal hypoglycemia episodes compared with NPH insulin in this population. These clinical benefits justify the use of basal insulin analogues, despite their high cost to treat patients with T2DM and CKD stages 3 and 4. Clinical Trials identifier: NCT02451917.