Background: LILRB2 is primarily expressed on myeloid cells and provides negative feedback during inflammatory response. A blocking antibody targeting LILRB2 in myeloid cells is in clinical trials. Various solid tumors are also enriched with these receptors. Here we investigate differences between LILRB2 expression in the local versus metastatic setting, influences on the tumor microenvironment, and effects on clinical outcomes for a group of solid tumors. Methods: Liver cancer (LC, N=532), urothelial carcinoma (UC, N=4125), pancreatic cancer (PC, N=5488), prostate adenocarcinoma (PA, N=5500) and non-small cell lung cancer (NSCLC, N=21604) tumors were tested at Caris Life Sciences (Phoenix, AZ) with NextGen sequencing of DNA (592-gene or whole exome) and RNA (whole transcriptome). Primary and metastatic sites were defined based on the biopsy site relative to known primary site. LILRB2-High (H) and -Low (L) expression was defined as top and bottom quartile of LILRB2 transcripts/million, respectively. PD-L1 (SP142; Positive (+): ≥2, ≥%5) expression was tested by IHC. Gene expression profiles were analyzed for transcriptomic signatures predictive of response to immunotherapy (T-cell inflamed score). Immune cell fractions were estimated with RNA deconvolution using quanTIseq. Mann-Whitney U and χ² tests were applied as appropriate with P-values adjusted for multiple comparisons. Overall survival data was obtained from insurance claims, and Kaplan-Meier estimates were calculated for molecularly defined subpopulations of patients. Results: Median expression of LILRB2 was increased in metastatic versus primary UC (1.23-fold, p<0.001), while similar expression was observed in PC (1.1, p=0.02), LC (0.87, p=0.68), PA (1.0, p=0.69), and NSCLC (0.96, p=0.003). Compared to LILRB2^L, LILRB2^H tumors had an increased prevalence of TP53 mutations in LC (LILRB2^H 53 vs 22%, p<0.05) and UC (64 vs 53, p<0.05). LILRB2^H tumors had an increased rate of PD-L1+ in PC (20 vs 11%), NSCLC (32 vs 11), and UC (40 vs 11, all p<0.05). Across all tumors, LILRB2^H have an increased amount of M1 (UC: 6.34-fold, PC: 1.50, LC: 3.30, PA: -H 0.5% vs -L 0.0, NSCLC: 2.08) and M2 macrophages (UC: 1.40, PC: 1.78, LC: 1.61, PA: 1.07, NSCLC: 1.47) as compared to LILRB2^L (p<0.05, all). LILRB2^H tumors were more frequently T cell-inflamed across investigated cancers (NSCLC: 66.4 vs 2.8%, PC: 60 vs 4.3, LC: 61 vs 3.0, PA: 69 vs 4.8, UC: 70 vs 2.6, all p<0.05). LILRB2^H PC tumors had decreased overall survival (HR 1.21 [1.0-1.2]). In NSCLC, LILRB2^H tumors had longer time on pembrolizumab (HR 0.87 [0.78-0.97]). Conclusions: LILRB2 expression was associated with biomarkers of response to immunotherapy such as PD-L1+ and an increased proportion of T cell-inflamed tumors. These data suggest that PC, NSCLC and UC tumors could benefit from a combination of ICI and LILRB2-blocking antibodies.