A mathematical approach for evaluating the error associated with sampling chromatographic peaks is developed and evaluated. Simulated chromatographic peaks having selected degrees of exponentially modified character and signal-to-noise ratios (S/N) are numerically generated. Total area is calculated for each peak by summing all data points representing the peak. Finite sampling rates are simulated by summing every nth point, where n corresponds to a sampling interval, then by applying the Newton-Cotes rule to obtain an approximation to the total area. The approximation is compared to the total area to obtain an error attributable to finite sampling. The number of data points between samples is kept large so that the total area is a good approximation to the true area of the peak. General trends in the results indicate that the sampling error decreases as S/N increases and, for a given sampling rate, the sampling error decreases as the peak becomes more skewed. The sampling error has a nearly exponential dependence on the sampling interval.