Although food has been proposed as a possible factor on population regulation of small mammals, evidence of food restriction in field conditions is still lacking. Food restriction is generally thought to occur in high-density years of oscillating populations of small mammals. By using two body condition indices (muscle protein concentration and cleaned carcass index), we investigated variations of body condition of the greater long-tailed hamster (Tscherskia triton) and the striped hamster (Cricetulus barabensis) in the Raoyang County of the North China Plain during a population decline- and low-phase from 1998 to 2003. Our laboratory experiments showed that food restriction reduced muscle protein concentration and cleaned carcass index of greater long-tailed hamster. We found there were significant differences in body condition of hamsters over the six years. Both the muscle protein concentration and cleaned carcass index of T. triton were significantly and positively correlated with its yearly trap successes, which reflected yearly population densities. Muscle protein concentration of C. barabensis was positively correlated with the total yield of the wheat. This observation supported the hypothesis on the in-phase relation between food resources or availability and rodent abundances. During the four seasons, muscle protein concentration of two species was highest in autumn, and lowest in summer. These two body condition indices of wild T. triton in summer and in low phases were similar to 60% food restricted animals raised in the laboratory, but significantly lower than nonfood- restricted animals. Our results indicated that malnutrition occurred in summer when food was not abundant, and in low density years with low grain production (an index reflecting food abundance of rodents) in our study region. Therefore, we concluded that food limitation may play an important role in seasonal and inter-annual population fluctuation of these two hamster species.