BackgroundThe high incidence of post-stroke depression (PSD) during rehabilitation exerts a negative effect on the treatment and functional recovery of patients with stroke and increases the risk of mortality. It is necessary to screen PSD in the rehabilitation stage and thus provide effective intervention strategies. However, existing measurements used to assess PSD in the rehabilitation stage in patients with stroke lack specificity. This study aimed to develop a clinical measure to assess symptoms of PSD in the rehabilitation stage.MethodsThe research team created the initial items through a literature review and semi-structured interviews of patients with stroke. Then, the symptom-related items were estimated by three panels: healthcare professionals (N = 41), Delphi experts (N = 15), and patients with stroke in the rehabilitation stage (N = 30).ResultsThe literature review and semi-structured interview produced 51 symptom-related items including six domains, and the items were reduced to 47 by the healthcare professionals. The symptom-related items were further reduced to 33 items by a two-round Delphi consultation. The initiative coefficients of the two Delphi rounds were 71.4 and 100%, the expert authority coefficients were both 0.85, Kendall’s W were 0.152 and 0.408 (p < 0.01), and the coefficient of variation (CV) were 0.05–0.32 and 0.00–0.18, respectively. The item-level content validity index (I-CVI) was 0.53–1.00, the scale-level CVI/universal agreement (S-CVI/UA) was 0.26, and the S-CVI/average (S -CVI/Ave) was 0.85 for the first found Delphi consultation; the I-CVI was 0.67–1.00, the S-CVI/UA was 0.61, and the S-CVI/Ave was 0.97 for the second round Delphi consultation. All content validity indicators have been significantly improved compared with the first round. Using mean ≥ 4 and full score ≥ 0.5, combined with CV ≤ 0.16 as the item criteria, a clinical measure of PSD with 33 items and 6 dimensions (cognition, sleep, behavior, emotion, body, and guilt) was finally formed after two rounds. The patients with stroke made no further revisions after evaluation.ConclusionThe research team developed a specific tool with good content validity to assess the symptoms of PSD in the rehabilitation stage.