Background: Climate change and pollution are among the five most important threats posed to biodiversity (Parmesan et al. 2022). Besides, it is now admitted that the unprecedented levels of environmental stress undergone by biodiversity lead to time scale overlaps between ecological and evolutionary responses (Carroll et al. 2007). As chemical pollutants constitute potent sources of environmental stress, their ability to trigger micro-evolutionary processes in exposed populations is increasingly studied. Understanding the evolutionary impact of these pollutants is also a new challenge for ecological risk assessment, especially in the context of global change where populations are often exposed to complex cocktails of contaminants along with other anthropogenic and natural stressors, which may interact in various ways. Therefore, it has become critical to assess chemicals’ long-term ecological risk under various climatic scenarios. A first step towards this end is to evaluate the genetic variability in stress responses, upon which selection might operate. In the present study, we addressed this issue by assessing the between-population variability in transcriptomic responses of a model invertebrate to trace metallic elements (as widespread persistant contaminants) and warming. We chose copper, as an essential element that becomes toxic at high doses, which is widespread and used as pesticide even in organic agriculture, and lithium, as an emergent contaminant associated with the development of digital technologies). Materials and methods: To this end, seven day- old F1s obtained in the laboratory from adults collected in five populations or strains of the great pondsnail Lymnaea stagnalis, were exposed for three weeks (juvenile period) to a sublethal concentration of lithium (100 µg/L) and copper (20 µg/L), separately or as a mixture, under two temperature regimes (20°C, considered as optimal under local conditions, and 24°C, following a +4° warming scenario). The effects of metals and temperature were estimated on transcriptomic expression (differential expression relative to controls, RNAseq), together with snail growth and cardiac frequency, and compared across populations. Results: Growth was faster at 24°C than at 20°C, reduced under copper exposure, and not affected by lithium. The two factors interacted significantly with the factor population, indicating genotype x environment interactions at the level of populations. Cardiac activity was negatively related to age and size, and this relation was steeper at 24°C, especially under lithium and control conditions (Temperature by Metal interaction). Population had not effect alone nor interacted with other factors on this trait. Transcriptomic analysis (Deseq2) showed a main effect of temperature, with 711 differentially expressed genes (DEGs). Among these, only 68 DEGs were shared by the five population. Lithium had almost no detectable effect on the transcriptome, whereas copper exposure induced substantial differential expression (538 DEGs), with no gene found in common among the five populations. While these results reflect a strong influence of the genetic background on the response of L. stagnalis to the tested factors, the functional significance of this between- population component of transcriptomic expression is currently explored through annotation (with a particular attention to regulatory factors) and enrichment analyses, which will be presented at the conference. Conclusions: The present results suggest that a species ability to cope with thermal and metal stress, as well as their combination, depends strongly on the population genetic background. Such interactions indicate that current ecological risk assessment of chemicals, which ignore warming and genetic influences, may be underprotective.