Catecholamines are a class of biogenic amines that play an important role as neurotransmitters and hormones. We developed and validated a rapid, specific and sensitive LC-MS/MS method for quantitative determination of catecholamines in human urine. Linearity, specificity, sensitivity, precision, accuracy, matrix effect, carryover, analyte stability, method comparison and reference range were evaluated. The catecholamine measurements were not affected by 35 structurally-related drugs and metabolites. The outstanding specificity was achieved by use of a specific diphenylborate-based solid phase extraction and subsequent selective LC-MS/MS analysis. Excellent sensitivity, accuracy and precision (average intra-assay variations <2.9 % and inter-assay variations <4.6 %) were obtained. The method was successfully applied in the study of day-to-day biological within- and between-subject variations of 25 healthy people under free-living conditions over three consecutive days. We observed that catecholamine excretions for second morning sampling had least day-to-day within-subject variation and excellent reproducibility. This work is one of the rare studies on these topics and represents the first utilization of advanced LC-MS/MS technology. Additionally, we found significant correlations between spot and conventional 24 h collections of human urine (n = 22, r > 0.853, p < 0.0001). These findings suggest that determining the catecholamine concentrations in the second morning urine sample presents accurate, convenient and reliable measurement of catecholamine excretions. In addition, consistent and significant diurnal variations for norepinephrine and epinephrine excretions were observed during the three-day period, while dopamine did not exhibit a diurnal rhythm. The LC-MS/MS method presented here is rapid, sensitive and specific, which could be an advantage in clinical laboratories.