AbstractWind erosion is an important soil degradation process in the semi‐arid Pampas of Argentina, but no attempts have been made to predict the process in this region. One limitation for the use of event‐based wind erosion prediction models is the lack of reliable climatic data to initiate these models. As an effort to apply wind erosion models, we compared field wind erosion measurements carried out during 4 years in a bare soil reference plot (RP) and during 3 years in the same soil with different tillage conditions: conventional (CT) and no‐till (NT). Wind erosion was predicted with: (1) subroutines for single storm event versions of the wind erosion prediction system (WEPS) and the revised wind erosion equation (RWEQ), for which the climatic data of each single storm were used, and (2) the wind erosion equation (WEQ) for full rotation periods, for which long‐term climatic records (1961–2004) were used. Wind erosion field measurements were carried out with BSNE samplers in 1‐ha plots. Regression of predicted soil erosion rates from WEPS (y = 0·5192x + 0·0589, R2 = 0·89) and RWEQ (y = 0·5691x − 7·071, R2 = 0·90) predictions with field measurements of soil erosion rates obtained in RP (54·51 kg m−1 on average) were highly significant, but both models underestimated wind erosion by 40–45 per cent. Predictions of wind erosion made with RWEQ were highly sensitive to variations in the soil crusting factor (SCF), varying from 60·5 t m−1 when predictions for a single storm were made using the SCF default data to 0·699 t m−1 when SCF was deduced from visual field observations. The WEQ predictions agreed adequately with measured erosion for 16 rotation periods either when using a climatic C factor value of 92, corresponding to the 1961–2004 period (y = 0·9422x − 1·9248, R2 = 0·96) or a C factor value of 80, corresponding to the moister 1985–2004 period (y = 0·7612x − 1·5543, R2 = 0·96). Neither WEPS nor RWEQ predicted the low amounts measured in CT and NT (3·86 kg m−1 on average) for storms lasting approximately 24 hours. High plant or residue soil coverage as well as high oriented surface roughness eliminated erosion according to WEPS and RWEQ. These results indicated that WEQ can be used as a reliable prediction model for long‐term predictions of wind erosion in the semi‐arid Pampas, even when run with limited available climatic data for this region. Copyright © 2008 John Wiley & Sons, Ltd.