Abstract: The measurement uncertainty for the analytical results of cyanide in groundwater samples by isonicotinic acid-pyrazolone spectrophotometric method was evaluated using continuous propagation model of uncertainty. The main sources of measurement uncertainties were from sub-uncertainties of sample preparation, calibration solution preparation, calibration curve fitting and instrument measurements, etc. The double error regression method was used for calibration curve fitting. The results showed that the lower the concentration of cyanide in the sample, the greater the relative uncertainty of the measurement results.