In this study, we aimed to explore the impact of exacerbation history on future exacerbation and mortality with different inhaled drugs in chronic obstructive pulmonary disease (COPD) patients based on a Global Initiative Chronic Obstructive Lung Disease (GOLD) A and B classifications. This observational study was based on the cohort study Real World Research of Diagnosis and Treatment of COPD (RealDTC). We collected data from COPD patients in China from 1 July 2017 to 31 December 2022. Patients were followed up until December 2023 or death. Further, we separated GOLD A and B patients into GOLD A0 and B0, who had no exacerbation during the previous year, and GOLD A1 and B1, who had only one exacerbation during the previous year. Study outcomes included moderate-to-severe exacerbation, hospitalisation, frequent exacerbation in the first year and all-cause mortality during total follow-up. Of the 8318 eligible patients, GOLD E group of patients suffered from a greater risk of exacerbation in the first year and death than patients in the GOLD A and B groups. GOLD A1 group had a higher risk of moderate-to-severe exacerbation (hazard ratio (HR) = 2.087; 95% confidence interval (CI) = 1.419-3.068), hospitalisation (HR = 1.704; 95% CI = 1.010-2.705) and frequent exacerbation (HR = 1.983; 95% CI = 1.046-3.709) compared to GOLD A0. GOLD B1 group had a risk of moderate-to-severe exacerbation (HR = 1.321; 95% CI = 1.105-1.679) and mortality (HR = 1.362; 95% CI = 1.026-1.963) that exceeded the risk in GOLD B0 group. The treatment outcome of different inhaled drugs had no statistical differences in GOLD A0 group. In GOLD A1 group, only inhaled corticosteroids (ICS), in addition to long-acting β-2 agonist (LABA) and long-acting muscarinic antagonist (LAMA), reduced the risk of moderate-to-severe exacerbation in the first year compared to only LAMA. As for the GOLD B0 group, LABA and LAMA decreased the odds of moderate-to-severe exacerbation, hospitalisation, frequent exacerbation and mortality compared to only LAMA. ICS, LABA, and LAMA in GOLD B0 also down-regulated the risk of frequent exacerbation, compared to only LAMA. In addition, GOLD B1 patients treated with LABA and LAMA or ICS, LABA, and LAMA had a lower risk of moderate-to-severe exacerbation and hospitalisation. Meanwhile, ICS, LABA, and LAMA also reduced the risk of frequent exacerbation and mortality, compared to only LAMA in the multivariate Cox analysis. Compared to the GOLD A or B group without exacerbation history, GOLD A patients with exacerbation history had a higher risk of future exacerbation, and GOLD B patients with exacerbation history had a higher risk of future exacerbation and mortality and benefited more from triple inhaler therapy.