Physical diseases co-occur with late-life depression (LLD). The influence of physical diseases and the subjective perception of physical health (PPH) on treatment outcome in LLD, however, is not well understood. To assess the association of physical diseases and PPH with the outcomes of 2 different types of psychotherapy in LLD. This post hoc secondary analysis of a multicenter, observer-blinded, controlled, parallel-group randomized clinical trial assessed participants 60 years or older with moderate to severe depression recruited at 7 psychiatric-psychotherapeutic outpatient trial sites in Germany from October 1, 2018, to November 11, 2020. Data analysis was performed from April 1 to October 31, 2023. Patients received LLD-specific cognitive behavioral therapy (LLD-CBT) or supportive unspecific intervention (SUI). Depression severity, response, and remission were measured during treatment and at 6-month follow-up by the change in the 30-item Geriatric Depression Scale (GDS) score. Physical health and PPH were assessed by the number of physical diseases, Charlson Comorbidity Index (CCI), and the World Health Organization Quality of Life Brief Version physical health subscale. A total of 251 patients were randomized to LLD-CBT (n = 126) or SUI (n = 125), of whom 229 (mean [SD] age, 70.2 [7.1] years; 151 [66%] female) were included in the intention-to-treat analysis. Patients with low and moderate PPH at baseline had significantly less reduction in the GDS score across both treatment groups than patients with high PPH (estimated marginal mean difference [EMMD], 2.67; 95% CI, 0.37-4.97; P = .02 for low PPH and EMMD, 1.82; 95% CI, 0.22-3.42; P = .03 for moderate vs high PPH). Higher PPH at baseline was associated with higher likelihood of response (odds ratio [OR], 1.04; 95% CI, 1.00-1.06; P = .009) and remission at the end of treatment (OR, 1.04; 95% CI, 1.02-1.08; P = .002) and response (OR, 1.05; 95% CI, 1.02-1.08; P < .001) and remission at follow-up (OR, 1.06; 95% CI, 1.03-1.10; P < .001) across both treatment groups. However, a significant interaction of PPH with treatment group was observed with low PPH at baseline being associated with significantly larger reduction in GDS scores in SUI compared with LLD-CBT at the end of treatment (EMMD, -6.48; 95% CI, -11.31 to -1.64; P = .009) and follow-up (EMMD, -6.49; 95% CI, -11.51 to -1.47; P = .01). In contrast, patients with high PPH at baseline had a significantly greater reduction in GDS scores in LLD-CBT compared with SUI at all time points (week 5: EMMD, -4.08; 95% CI, -6.49 to -1.67; P = .001; end-of-treatment: EMMD, -3.67; 95% CI, -6.72 to -0.61; P = .02; and follow-up: EMMD, -3.57; 95% CI, -6.63 to -0.51; P = .02). The number of physical diseases or CCI at baseline did not have an effect on the change in GDS score, response, or remission, neither across both groups nor within either group. In this secondary analysis of a randomized clinical trial, subjective PPH was associated with treatment outcome, response, and remission in psychotherapy of LLD. Patients with LLD responded differently to LLD-CBT and SUI, depending on their baseline PPH score. Treatment approaches for patients with LLD should address PPH in personalized interventions. ClinicalTrials.gov Identifier: NCT03735576; Deutsches Register Klinischer Studien Identifier: DRKS00013769.