The Gravity Recovery and Climate Experiment (GRACE) satellite mission will provide new measurements of Earth's static and time-variable gravity fields with monthly resolution. The temporal effects due to ocean tides and atmospheric mass redistribution are assumed known and could be removed using current models. In this study we quantify the aliasing effects on monthly mean GRACE gravity estimates due to errors in models for ocean tides and atmosphere and due to ground surface water mass variation. Our results are based on simulations of GRACE recovery of monthly gravity solution complete to degree and order 120 in the presence of the respective model errors and temporal aliasing effects. For ocean tides we find that a model error in S2 causes errors 3 times larger than the measurement noise at n < 15 in the monthly gravity solution. Errors in K1, O1, and M2 can be reduced to below the measurement noise level by monthly averaging. For the atmosphere, model errors alias the solution at the measurement noise level. The errors corrupt recovered coefficients and introduce 30% more error in the global monthly geoid estimates up to maximum degree 120. Assuming daily CDAS-1 data for continental surface water mass redistribution, the analysis indicates that the daily soil moisture and snow depth variations with respect to their monthly mean produce a systematic error as large as the measurement noise over the continental regions.