Cell-cell interaction (CCI) analyses are becoming an indispensable discovery tool for cutting-edge single cell and spatial omics technologies, identifying ligand-receptor (LR) targets in intercellular communications at the molecular, cellular, and microenvironment levels. Different transcriptional-based modalities can add complementary information and provide independent validation of a CCI, but so far no robust methods exist to integrate CCI results together. To address this, we have developed a statistical and computational pipeline, Multimodal CCI (MMCCI), implemented in an open-source Python package, which integrates, analyzes, and visualizes multiple LR-cell-type CCI networks between multiple samples of a single transcriptomic-based modality as well as between multiple modalities. MMCCI implements new and in-depth downstream analyses, including comparison between biological conditions, network and interaction clustering, sender-receiver interaction querying, and pathway analysis. We applied MMCCI to integrate CCIs in our spatial transcriptomics datasets of aging mouse brains (from 10X Visium and BGI STOmics) and melanoma (10X Visium, 10X Xenium and NanoString CosMx) and identified biologically meaningful interactions. Using simulated data, we applied MMCCI integration on four popular CCI algorithms, stLearn, CellChat, Squidpy, and NATMI, and detected highly confident interactions while reducing false interaction discoveries through the integration of multiple samples. With MMCCI, the community will have access to a valuable tool for harnessing the power of multimodal single cell and spatial transcriptomics. MMCCI source code and documentation are available at:https://github.com/BiomedicalMachineLearning/MultimodalCCI.