Summary: Baseline serum PINP value was significantly and independently associated with the increased bone mineral density (≥ 3%) in both total hip and femoral necks by 12 months of romosozumab treatment in patients with treatment-naive postmenopausal osteoporosis. Purpose: Some patients fail to obtain a sufficiently increased hip bone mineral density (BMD) by romosozumab (ROMO) treatment. This study aimed to investigate the prognostic factor for increased hip BMD with ROMO in patients with treatment-naive postmenopausal osteoporosis. Methods: This prospective, observational, and multicenter study included patients (n = 63: mean age, 72.6 years; T-scores of the lumbar spine [LS], − 3.3; total hip [TH], − 2.6; femoral neck [FN], − 3.3; serum type I procollagen N-terminal propeptide [PINP], 68.5 µg/L) treated by ROMO for 12 months. BMD and serum bone turnover markers were evaluated at each time point. A responder analysis was performed to assess the patient percentage, and both univariate and multivariate analyses were performed to investigate the factors associated with clinically significant increased BMD (≥ 3%) in both TH and FN. Results: Percentage changes of BMD from baseline in the LS, TH, and FN areas were 17.5%, 4.9%, and 4.3%, respectively. In LS, 96.8% of patients achieved ≥ 6% increased LS-BMD, although 57.1% could not achieve ≥ 3% increased BMD in either TH or FN. Multiple regression analysis revealed that only the baseline PINP value was significantly and independently associated with ≥ 3% increased BMD in both TH and FN (p = 0.019, 95% confidence interval = 1.006–1.054). The optimal cut-off PINP value was 53.7 µg/L with 54.3% sensitivity and 92.3% specificity (area under the curve = 0.752). Conclusions: In a real-world setting, baseline PINP value was associated with the increased BMD of TH and FN by ROMO treatment in treatment-naive patients. [ABSTRACT FROM AUTHOR]