Long non-coding RNAs (lncRNAs) constitute a fascinating class of regulatory RNAs, widely distributed in eukaryotes. In plants, they exhibit features such as tissue-specific expression, spatiotemporal regulation, and responsiveness to stress, suggesting their involvement in specific biological processes. Although an increasing number of studies support the regulatory role of lncRNAs in model plants, our knowledge about these transcripts in relevant crops is limited. In this study we employ a custom pipeline on a dataset of over 1,000 RNA-seq studies across nine representative species of the family Cucurbitaceae to predict 91,209 non-redundant lncRNAs. LncRNAs were predicted according to three confidence levels and classified into intergenic, natural antisense, intronic, and sense overlapping. Predicted lncRNAs have lower expression levels compared to protein-coding genes but a more specific behavior when considering plant tissues, developmental stages, and response to stress, emphasizing their potential roles in regulating various aspects of plant-biology. The evolutionary analysis indicates higher positional conservation than sequence conservation, which may be linked to the presence of conserved modular motifs within syntenic lncRNAs. In short, this research provides a comprehensive map of lncRNAs in the agriculturally relevant Cucurbitaceae family, offering a valuable resource for future investigations in crop improvement.