AimsThe importance of programmed cell death ligand 1 (PD‐L1) expression has emerged in clinical trials of PD‐L1 target therapy in renal cell carcinoma (RCC). This study compares PD‐L1 assays in RCC.Methods and resultsTwo US Food and Drug Administration‐approved PD‐L1 assays (22C3 and SP142) and one research‐use only antibody (E1L3N) were used in a retrospective cohort of 591 patients with RCC. PD‐L1 positivity on tumour cells (TCs) and immune cells (ICs) and combined positive score (CPS) were evaluated. With the 22C3, SP142 and E1L3N assays, positive PD‐L1 expression on TCs ≥1% was observed in 24 (4.1%), 12 (2.0%) and 16 (2.7%) cases and on ICs ≥1% was observed in 132 (22.3%), 120 (20.3%) and 65 (11.0%) cases, respectively. PD‐L1 expression scores among the three assays showed moderate–high positive correlation (ρ = 0.599–0.835, P < 0.001). Assays appeared similar, although staining in ICs was comparatively less frequent with E1L3N. 22C3 showed frequent positivity in TCs. PD‐L1 expression on TCs was associated with papillary type 2 RCC (P < 0.001). IC infiltration and PD‐L1 expression on ICs were predominantly found in clear cell and papillary type 1 RCC (P < 0.05).ConclusionsProgrammed death 1 (PD‐1)/PD‐L1 target therapy may be beneficial for patients with papillary type 2 RCC, even if they are categorised as a heterogeneous group. PD‐L1 assays should be carefully selected, and accurate histological subtyping of RCC is needed prior to decisions on PD‐L1 testing, because of the different PD‐L1 expression observed among varying RCC subtypes.