Foraging is a fundamental behavior to ensure animals’ survival and reproduction. Foraging decisions should balance the risks of starvation and predation which usually varied both spatially and temporally in the wild. The energetic state-risk allocation hypothesis presented that the optimal trade-off between energetic gain from foraging and avoidance of predation should depend on animals’ current physiological states (e.g., hunger), on current environmental states (e.g. actual predation risk, food availability), and perhaps also on past and expected future states. However, in terms of the model of energetic state and predation risk allocation, studies in rodents have yielded mixed results. Therefore, we used a Y-shaped maze by a 10-min behavioral test to measure the discrimination to the predator odor of a Steppe Polecat (Mustela eversmannii) or no-predator (i.e. horse) odor in female Mongolian Gerbils (Meriones unguiculatus) deprived food for 18 h in field laboratory conditions. Then we investigated the feeding and antipredator efforts of the fasted female gerbils in the presence of predator odor with food or those of in the control no-predator odor with food at a neutral arena. We aimed to test the hypothesis that acute predation risk inhibited the foraging effort of the starving gerbils, and then to address the trade-off between the starvation and predation risk in the foraging decision of gerbils, as a desert rodent. We found that both the frequency and the duration of visiting the predator odor selection box was significantly less than that of visiting the control box (Wilcoxon signed ranks test: frequency, Z = 2.405, P = 0.016; duration, Z = 2.803, P = 0.005; Table 1), which indicated a remarkable aversive or fear response to the steppe polecat odors in female Mongolian gerbils. Furthermore, we observed that starving gerbils displayed their foraging effort both in the treatment with a predation risk and without predation risk, e.g. unchanged their foraging frequency compared to under the predation risk and under the control situation (Mann-Whitney test: Z = 1.514, P = 0.130; Fig. 2). It was just that，compared with the foraging latency (58.6 ± 35.5 s) under control situation，the starving gerbils under predation risks increased the foraging latency (110.4 ± 57.9 s), delaying their foraging (Mann-Whitney test: Z = 2.068, P = 0.039; Fig. 1), yet decreased the duration per foraging (Mann-Whitney test: Z = 2.856, P = 0.004; Fig. 1). Additionally, we noted that gerbils were higher in the effort of the up-investigating to the space for vigilance (Mann-Whitney test: frequency: Z = 3.187, P = 0.001; Fig.2; duration, Z = 2.725, P = 0.006; Fig. 3) and self-grooming for releasing fear in the treatment with the predation odor risk (Mann-Whitney test: Z = 2.289, P = 0.022, Fig. 2). Our results, thus, revealed that the acute predation risk could not completely inhibit the foraging activities in starving female Mongolian gerbils, and partly supported the energetic state-predation risk allocation hypothesis. These results suggested that the gerbil could respond to the hazard of starvation and predation by risk-adjusting or balancing, and trade off the energetic gain from foraging and cost of predation maximizing the current or lifetime fitness, which benefited to the Mongolian gerbils adapt to desert environment characterized by largely unpredictable food resources and relative higher predation-risk.