Selection by herbicides has resulted in widespread evolution of herbicide resistance in agricultural weeds. In California, resistance to glyphosate was first confirmed in rigid ryegrass in 1998. Objectives of this study were to determine the current distribution and level of glyphosate resistance in Italian ryegrass, and to assess whether resistance could be due to an altered target site. Seeds were sampled from 118 populations and seedlings were treated with glyphosate at 866 g ae ha−1. Percentage of survivors ranged from 5 to 95% in 54 populations. All plants from 64 populations died. One susceptible (S) population, four putatively resistant (R) populations, and one S accession from Oregon were used for pot dose–response experiments, shikimic acid analyses, and DNA sequencing. Seedlings were treated with glyphosate at eight rates, ranging from 108 to 13,856 g ae ha−1. Shoot biomass was evaluated 3 wk after treatment and fit to a log-logistic regression equation. On the basis of GR50 (herbicide rate required to reduce growth by 50%) values, seedlings from putatively R populations were roughly two to 15 times more resistant to glyphosate than S plants. Shikimic acid accumulation was similar in all plants before glyphosate treatment, but at 4 and 7 DAT, S plants from California and Oregon accumulated approximately two and three times more shikimic acid, respectively, than R plants. Sequencing of a cDNA fragment of the EPSPS coding region revealed two different codons, both of which encode proline at amino acid position 106 in S individuals. In contrast, all R plants sequenced exhibited missense mutations at site 106. Plants from one population revealed a mutation resulting in a proline to serine substitution. Plants from three R populations exhibited a mutation corresponding to replacement of proline with alanine. Our results indicate that glyphosate resistance is widespread in Italian ryegrass populations of California, and that resistance is likely due to an altered target enzyme.