The problem of crop acreage estimation using satellite data is discussed. The bias and variance of a crop estimate in an area segment obtained from the classification of its multispectral sensor data are derived as functions of the means, variances, and covariances of error rates, Linear discriminant analysis and class proportion estimation for the two-class case have been extended to include a third class of measurement units, where these units are mixed on the ground. Special attention is given to the investigation of mislabeling in training samples and its effect on crop proportion estimation. It is shown that the bias and variance of the estimate of a crop acreage proportion increase as the disparity in mislabeling rates of the two classes increases. The effect of mixed pixels on the bias and variance of the proportion estimate is also investigated.