Field experiments were conducted to determine if seeds would be produced on Triticum aestivum by Aegilops cylindrica hybrids in the field and, if it were, to determine the viability of the seeds produced. One, five, or 10 hybrids were planted into varying proportions of A. cylindrica and T. aestivum in a replacement series design. Percent seed set ranged from 0 to 5.5% in 1996 and from 0 to 9.2% in 1997. Seeds were set in all treatments. The average seed set was 2.3% in 1996 and 3.8% in 1997. No differences in seed set were found among treatments. The seeds produced were separated according to seed condition, either full or shriveled, and tested for germination. The germination of the seeds produced on the hybrids was not significantly different between years. The average germination for full seeds was 94% in both years and 79 and 84% for shriveled seeds in 1995 and 1996, respectively. Greenhouse studies were conducted to evaluate the rate of self-fertility of the BC1 generation and to identify morphological characteristics that could be used to identify the probable pollen donor parent and to predict the occurrence of seed set. In 1997 4.1% and in 1998 2.1% of BC1 plants set seeds. The average seed set was 0.3% in 1997 and 0.06% in 1998. It was not possible, using any morphological characteristic measured, to determine the identity of the parent serving as the pollen donor in the previous generation or to predict the occurrence of seed set in the BC1 generation. This is the first reported study to show that hybrids between T. aestivum and A. cylindrica have the ability, although limited, to backcross under field conditions and set seeds. Furthermore, the seeds produced are viable and will germinate and produce plants. With the millions of hectares of T. aestivum infested with A. cylindrica, even the limited ability to backcross is of concern for the movement of a herbicide-resistance gene.