Drugs targeting β-cells have provided new options in the management of T2DM; however, their role in β-cell regeneration remains elusive. The recent emergence of cell-based therapies such as autologous bone marrow-derived mesenchymal stem cells (ABM-MSCs) and mononuclear cells (ABM-MNCs) seems to offer a pragmatic approach to augment β-cell function/mass. This study aims to examine the efficacy and safety of ABM-MSC and ABM-MNC transplantation in T2DM and explores alterations in glucose–insulin homeostasis by metabolic studies. Thirty patients of T2DM with duration of disease ≥5 years, receiving triple oral antidiabetic drugs along with insulin (≥0.4 IU/Kg/day) with HbA1c ≤7.5%(≤58.0 mmol/mol), were randomized to receive ABM-MSCs or ABM-MNCs through targeted approach and a sham procedure (n = 10 each). The primary endpoint was a reduction in insulin requirement by ≥50% from baseline, while maintaining HbA1c <7.0% (<53.0 mmol/mol) during 1-year follow-up. Six of 10 (60%) patients in both the ABM-MSC and ABM-MNC groups, but none in the control group, achieved the primary endpoint. At 12 months, there was a significant reduction in insulin requirement in ABM-MSC (P < 0.05) and ABM-MNC groups (P < 0.05), but not in controls (P = 0.447). There was a significant increase in second-phase C-peptide response during hyperglycemic clamp in the ABM-MNC (P < 0.05) group, whereas a significant improvement in insulin sensitivity index (P < 0.05) accompanied with an increase in insulin receptor substrate-1 gene expression was observed in the ABM-MSC group. In conclusion, both ABM-MSCs and ABM-MNCs result in sustained reduction in insulin doses in T2DM. Improvement in insulin sensitivity with MSCs and increase in C-peptide response with MNCs provide newer insights in cell-based therapies.