Our aim was to assess the impact of increase in homeostasis model assessment of insulin resistance (HOMA-IR) on the development of type 2 diabetes in Japanese individuals with impaired insulin secretion (IIS). This study included 2,209 participants aged 30–69 without diabetes at baseline who underwent comprehensive medical check-ups between April 2006 and March 2007 at Saku Central Hospital. Participants were classified into eight groups according to the combination of baseline IIS status (non-IIS and IIS) and category of HOMA-IR change between the baseline and follow-up examinations (decrease, no change/small increase, moderate increase, and large increase). Type 2 diabetes was determined from fasting and 2 h post-load plasma glucose concentrations at the follow-up examination between April 2009 and March 2011. At baseline, 669 individuals (30.3%) were classified as having IIS. At follow-up, 74 individuals developed type 2 diabetes. After adjusting for confounding factors including baseline HOMA-IR values, the multivariable-adjusted odds ratios (95% confidence intervals) for type 2 diabetes in the non-IIS with a decrease (mean change in HOMA-IR: −0.47), non-IIS with a moderate increase (mean change in HOMA-IR: 0.28), non-IIS with a large increase (mean change in HOMA-IR: 0.83), IIS with a decrease (mean change in HOMA-IR: −0.36), IIS with no change/small increase (mean change in HOMA-IR: 0.08), IIS with a moderate increase (mean change in HOMA-IR: 0.27), and IIS with a large increase (mean change in HOMA-IR: 0.73) groups, relative to the non-IIS with no change/small increase (mean change in HOMA-IR: 0.08) group were 0.23 (0.04, 1.11), 1.22 (0.26, 5.72), 2.01 (0.70, 6.46), 1.37 (0.32, 4.28), 3.60 (0.83, 15.57), 5.24 (1.34, 20.52), and 7.01 (1.75, 24.18), respectively. Moderate and large increases in HOMA-IR had a strong impact on the development of type 2 diabetes among individuals with IIS in this Japanese population.