Background: Blockade of the programmed death-1 (PD-1) receptor results in impressive activity in patients (pts) with metastatic melanoma (MM). Mechanisms of antitumor effect are not fully understood but may be associated with the degree of T cell receptor (TCR) repertoire clonality in tumor infiltrating lymphocytes (TIL). We hypothesized that differences among multiple tumors in the same patient may be reflected in heterogeneity of TCR repertoire clonality in TIL. We tested this hypothesis using TCR sequencing in pts with MM treated with pembrolizumab. Methods: Post-treatment samples were obtained from different lesions at the same time points. TCR repertoire quantification was performed using high-throughput sequencing of the rearranged TCR β-chain genes; these were amplified and sequenced using the survey ImmunoSeq assay in a multiplexed PCR method using primers to TCR Vβ and Jβ gene segments. Productive TCR sequences were used to obtain a clonality metric, and frequency of each T cell clone was obtained by comparing the number of reads generated by each unique CDR3 sequence. The proportion of TIL was determined using quantitative immunohistochemistry. Variations in proportion of TIL and clonality between samples were analyzed using ANOVA test. Tests were performed in duplicate. Results: Eight tumors from two pts receiving pembrolizumab (Four from pt A and four from pt B, obtained 2 and 4 months after treatment initiation, respectively) were analyzed. Significant intrapatient intertumoral heterogeneity in TCR repertoire was observed (pt A: p = 0.000153 and p = 0.013/pt B: p = 0.00083 and p = 9.01E-07 for differences in proportion of TIL and clonality, respectively). There was a high correlation between duplicates (r² = 0.86- > 0.99), suggesting these findings were not the result of experimental variation. Conclusions: Our data suggest significant intrapatient intertumoral heterogeneity in the proportion of TIL and clonality of the TCR repertoire at the same post-treatment timepoint. Caution should be exercised in interpreting the results of TCR analyses from single tumor biopsies in patients treated with checkpoint blockade.