The effect of protein-calorie malnutrition or nutritional recovery and chronic aerobic low intensity and volume training exercise, both separately and together, serum glucose, protein, albumin and hemoglobin were studied in female Fisher rats subjected to a 9-week program of swimming exercise (30 min/day, 5 days/week). After weaning (28 days old) fifty-two female Fisher rats were distributed into two groups: Control diet (AIN-93M; n = 24) and Low protein diet (5% protein; n = 28) during 30 days. Following half of the Low protein group were changed your diet to control diet and the three resulting groups were again assigned to a Sedentary or Exercise-trained subgroup. Animals were kept in individual cages and received water and food ad libitum. After 9 weeks the animals were anesthetized and sacrificed. Results were compared by two-way ANOVA (p ≤ 0.05). There was no significant difference between animals by the exercise effect, diet effect or interaction between both for serum albumin. There were significant differences between groups by a diet effect for serum hemoglobin (mmol/l) and glucose (mmol/l). The concentrations were respective higher in Control (26.42 ± 2.08); (7.95 ± 1.74) following by Recovery (22.74 ± 0.87); (6.85 ± 0.89) and the lower value was found in Malnutrition (15.52 ± 2.33); (6.67 ± 0.81) animals. Serum protein (g/l) was significant difference by a diet effect, and has been higher in Recovery, following by Control and the lower value was showed by Malnutrition animals. This parameter also differs by interaction between exercise and diet. Sedentary Recovery showed a higher value (71.36 ± 4.04) like the Trained Recovery (69.71 ± 3.68) and Trained Control (69.75 ± 3.75), following by Sedentary Control (63.63 ± 4.74) and the lowers values were found in Trained Malnutrition (60.90 ± 4.77) like the Sedentary Malnutrition (60.34 ± 5.25). The results show that exercise itself was not efficient to alter serum concentration of glucose, albumin and hemoglobin of rats in different nutritional status and indicate that low volume training is not capable of altering these parameters. Its need to be made new studies with different intensities and volume of training.