Purpose: To verify the reciprocal longitudinal relationships between cardiorespiratory fitness (CRF), percentage body fat (%body fat), and metabolic syndrome in Brazilian primary school students. Method: This longitudinal study involved 420 children and adolescents followed for 3 years (2011–2014). The continuous Metabolic Syndrome (cMetSyn) score was calculated by summing adjusted z scores of glucose, systolic blood pressure, total cholesterol/high-density lipoprotein cholesterol ratio, triglycerides, and waist circumference. The CRF was assessed using running/walking tests, and %body fat was assessed through sex-specific 2-site skinfold thickness. Cross-lagged panel models were used to analyze longitudinal reciprocal relationships between CRF and %body fat with cMetSyn. Results: Results indicated that 2011 %body fat significantly predicted both 2014 CRF scores and 2014 cMetSyn scores (P < .001); however, 2011 CRF only predicted 2014 %body fat (P < .001) but not 2014 cMetSyn (P = .103). Furthermore, 2011 cMetSyn predicted 2014 %body fat (P = .002). The model explained 36%, 48%, and 37% of the variance in 2014 CRF, %body fat, and cMetSyn, respectively. Conclusion: The results suggest a reciprocal inverse relationship between %body fat and metabolic syndrome risk and that %body fat may play a more important role in the risk of developing metabolic syndrome compared with CRF.