Genetic and ecological factors may interact in their effects on fitness. Such interactions are thus to be expected between inbreeding and exposure of a population to a toxicant. The magnitude of inbreeding depression is thought to increase in stressful environments. To test this hypothesis, we investigated the combined effects of environmental conditions and inbreeding on fitness in the self-fertile snail Lymnaea stagnalis, using a stress gradient (0–2) applied to a 100 isolated and paired lineages: laboratory control (0), outdoor microcosm control (1) and pesticide exposure under outdoor microcosm conditions (2). Outdoor stress conditions were maintained for 28 days prior to measurements of fitness traits (fecundity, hatching success, and size at hatching) under laboratory conditions, so that delayed environmental effects could be estimated. Under laboratory control conditions, we found significant initial family level heterogeneity for most measured traits, including physiological performances as assessed through energetic biomarkers. Whatever the environmental conditions, inbreeding depression was very low for progeny performances. Negative values of self-fertilization depression (SFD) were obtained. Unexpectedly, SFD showed a negative relationship with the assumed stress intensity, reflecting a higher sensitivity under pairing than under selfing, mostly due to parental fecundity. This suggests that stressful conditions may favour selfing. Stress intensity increased the distribution limits of both depression indices, suggesting that changes in fitness are less predictable in a population under stress. Implications of such findings for environmental risk assessment of pesticides are discussed.