To establish semen quality scores in a population of sperm donors and determin e whether the scores can be used to predict pregnancy rates after donor insemina tion. Retrospective study. Infertility clinic at a tertiary care teaching hospit al between 1993-2001. One hundred eleven women who underwent IUI with sperm fro m 27 anonymous donors. None. The semen analysis results before freezing and afte r thawing were analyzed, and overall donor insemination semen quality (DI-SQ) a nd donor insemination relative quality (DI-RQ) scores were calculated. After ad justing for female characteristics, the scores were compared with samples that d id and did not result in pregnancy. Of 111 patients, 70 had at least one pregnan cy, and 60 had at least one live birth, with a mean of 6.52 ±4.67 IUI cycles pe r patient. Five significant risk factors for low pregnancy and live birth rates were identified: female infertility factor, positive laparoscopy, older maternal age, low number of previous births, and lack of ovulatory stimulation. After ad justing for these factors, both prefreeze and postthaw DI-SQ scores were statis tically significantly associated with IUI live birth rates. Using only the sampl es with a DI-SQ score of>110 doubled the expected live birth rate,compared with usi ng samples with a DI-SQ score less than 100, from 8.5%to 16.1%. The DI-SQ sc ore was an effective predictor of pregnancy and live birth outcomes in IUI patie nts who underwent artificial insemination with anonymous donor semen. The DI-SQ score could also be used by sperm banks to help select donors. To establish semen quality scores in a population of sperm donors and determin e whether the scores can be used to predict pregnancy rates after donor insemina- tion. Retrospective study. Infertility clinic at a tertiary care teaching hospit al between 1993-2001. One hundred eleven women who underwent IUI with sperm fro m 27 anonymous donors. None. The semen analysis results before freezing and afte r thawing were analyzed, and overall donor insemination semen quality (DI-SQ) a nd donor insemination relative quality (DI-RQ) scores were calculated after After just with for samples that d id and did not result in pregnancy. 6.52 ± 4.67 IUI cycles pe r patient. Five significant risk factors for low pregnancy and live birth rates were identified: female infertility factor, positive laparoscopy, older maternal age, low number of previous births , and lack of ovulatory stimulation. After ad justing for these factors, both prefreeze and postthaw DI-SQ scores were statisically significantly associated with IUI live birth rates. Using only the sampl es with a DI-SQ score of> 110 doubled the expected live birth rate, compared with usi ng samples with a DI-SQ score less than 100, from 8.5% to 16.1%. The DI-SQ sc ore was an effective predictor of pregnancy and live birth outcomes in IUI patients who underwent artificial insemination with anonymous donor semen. The DI-SQ score could also be used by sperm banks to help select donors.