Many natural populations experience inbreeding and genetic drift as a consequence of nonrandom mating or low population size. Furthermore, they face environmental challenges that may interact synergistically with deleterious consequences of increased homozygosity and further decrease fitness. Most studies on inbreeding–environment (I-E) interactions use one or two stress levels, whereby the resolution of the possible stress and inbreeding depression interaction is low. Here we produced Drosophila melanogaster replicate populations, maintained at three different population sizes (10, 50 and a control size of 500) for 25 generations. A nutritional stress gradient was imposed on the replicate populations by exposing them to 11 different concentrations of yeast in the developmental medium. We assessed the consequences of nutritional stress by scoring egg-to-adult viability and body mass of emerged flies. We found: (1) unequivocal evidence for I-E interactions in egg-to-adult viability and to a lesser extent in dry body mass, with inbreeding depression being more severe under higher levels of nutritional stress; (2) a steeper increase in inbreeding depression for replicate populations of size 10 with increasing nutritional stress than for replicate populations of size 50; (3) a nonlinear norm of reaction between inbreeding depression and nutritional stress; and (4) a faster increase in number of lethal equivalents in replicate populations of size 10 compared with replicate populations of size 50 with increasing nutritional stress levels. Our data provide novel and strong evidence that deleterious fitness consequences of I-E interactions are more pronounced at higher nutritional stress and at higher inbreeding levels.