Previous studies have shown that breath ammonia (breath-NH3) concentration is associated with blood urea nitrogen (BUN) levels. However, interindividual variations in breath-NH3 concentrations were observed. Thus, the present study aimed to assess the effect of oral cavity conditions on breath-NH3 concentration and to validate whether the measurement of breath-NH3 concentration is feasible in clinical settings. A total of 125 individuals, including patients with stage 3 to 5 chronic kidney disease (CKD3-5), those on dialysis, and healthy participants, were recruited. A nanostructured sensor was used to detect breath-NH3 concentrations. Pre- and post-gargling as well as pre- and post-hemodialysis (HD) breath-NH3, salivary pH, and salivary urea levels were measured. Breath-NH3, salivary urea, salivary pH, and BUN levels were positively correlated to each other. Breath-NH3 concentrations were associated with BUN levels (r = 0.43, p < 0.001) and were significantly higher in CKD3-5 (p < 0.005) and dialysis patients (p < 0.001) than in healthy participants. Higher correlation coefficients were noted between breath-NH3 concentrations and BUN levels during follow-up (r = 0.59-0.94, p < 0.05). When the cutoff value of breath-NH3 was set at 523.65 ppb, its sensitivity and specificity in predicting CKD (BUN level >24 mg dl-1) were 87.6% and 80.9%, respectively. Breath-NH3 concentrations decreased after HD (p < 0.001) and immediately after gargling (p < 0.01). Breath-NH3 concentration, which was affected by gargling, was correlated to BUN level. The measurement of breath-NH3 concentration using the nanostructured device may be used as a tool for CKD detection and personalized point-of-care for CKD and dialysis patients. The current study had a small sample size. Thus, further studies with a larger cohort must be conducted to validate the effect of oral factors on breath-NH3 concentration and to validate the benefit of breath-NH3 measurement.