Abstract. GOCE will be the first satellite gravity mission measuring gravity gradients in space using a dedicated instrument called a gradiometer. High resolution gravity field recovery will be possible from these gradients. Such a recovery requires a proper description of the gravity gradient errors, where the a priori error model is for example based on end-to-end instrument simulations. One way to test the error model against real data, i.e. to see if the a priori model really describes the actual error, is to compare along track interpolated gradients with the measured gradients. The difference between the interpolated and measured gravity gradients is caused by, among others, the interpolation error and the measurement errors. The idea is that if the interpolation error is small enough, then the differences should be predicted reasonably well by the error model. This paper discusses a simulation study where the gravity gradient errors are generated with an end-to-end instrument simulator. The measurement error will be compared with the interpolation error and we will assess the latter as a function of the sampling interval.