Controlled environment and field research were undertaken to study the performance of selected Moroccan bread wheats (Triticum aestivum L.): 'Potam', 'Nasma' and 'Florelle', durum wheat (Triticum durum Desf.): 'Keyperounda' and 'Cocorit', and barleys (Hordeum vulgare L.): 'Arig 8' and 'ACSAD 60' under different moisture regimes. 1-(Malonylamino)cyclopropane-1-carboxylic acid (MACC) accumulation, grain yield stability, grain yield, and yield-component relationships, water-use efficiency, and water relationships were investigated. The wheat crop model, SIMTAG (Simulation of Triticum Aestivum Genotypes), was tested and evaluated against actual field data;Under controlled environments, Nasma, Keyperounda, Cocorit, and Arig 8 accumulated MACC. Stressed plants accumulated two to five times higher levels compared with nonstressed plants. Durum wheats and bread wheat accumulated the highest and lowest levels of MACC, respectively. Field studies conducted in Morocco from 1985 to 1987 using a line-source sprinkler irrigation system showed that Potam, Nasma and Florelle, Keyperounda, Cocorit, Arig 8 and ACSAD 60, behaved differently when they were subjected to six soil moisture regimes. Cultivar effects were predominant. Based on three indices, i.e., regression coefficient (b), regression intercept (a) and drought susceptibility index(s), it was inferred that Potam, Nasma and ACSAD 60 were more drought tolerant than the other cultivars. They maintained a yield advantage in drier environments by producing greater grain numbers m[superscript]-2 and a higher harvest index. Moreover, to increase yield and dry matter production water use had to be improved through improved water extraction from deep soil horizons. A maximum amount of water use should be oriented to crop transpiration by minimizing soil evaporation early in the season. Finally, the wheat model SIMTAG showed considerable promise as a tool in predicting wheat crop growth and yield responses. Despite discrepancies in green area, kernel weight, and spikes m[superscript]-2, the model predicted phenological stages (one to seven day deviation), soil moisture content (R[superscript]2 = 0.71), grain yield (R[superscript]2 = 0.67) and dry matter (R[superscript]2 = 0.87) reasonably well.