Minimizing the number of patient-reported outcome measures (PROMs) can reduce patient burden. The primary aim of the present study was to investigate whether physical therapists (PTs) can estimate psychological PROM scores in patients with low back pain (LBP) through physical therapy evaluation. The secondary aims were; 1) to investigate whether the clinical experiences of PTs influence correlations between PT estimates and psychological PROM scores, and 2) to investigate the sensitivity and specificity of PT estimates for the psychological features detected by the PROMs. We recruited hospitalized patients owing to LBP, who underwent evaluation by PTs on the initial day of hospitalization. Patients completed PROMs, including the Pain Catastrophizing Scale (PCS), Tampa Scale for Kinesiophobia, and Hospital Anxiety and Depression Scale immediately before the initial physical therapy session. PTs rated the magnitude of patient kinesiophobia, pain catastrophizing, anxiety, and depression using an 11-point numerical rating scale (NRS; 0 = not detected at all, 10 = very highly detected) through physical therapy evaluation immediately after the initial session. The PTs were blinded to the PROM results. We categorized PTs into two subgroups (PTs with ≥4 years and those with <4 years of clinical experience). Data from 78 patients (mean [SD] age = 60.5 [16.3] years) and 21 PTs were analyzed. A statistically significant but weak correlation (P = .04, Spearman's ρ = .24) was detected only in the total PCS scores and PT NRS scores in a dataset of all patients and PTs. Further, there were no statistically significant differences in correlations (all P >.05) between the two subgroups of PTs in all measures. Low sensitivity and high specificity of PT estimates for psychological features through physical therapy evaluation were identified in all PROMs when PT NRS scores were categorized into the binary score by 5 (negative: <5; positive: ≥5).