The Gonozyme test (Abbott Laboratories; North Chicago, IL), a solid-phase enzyme immunoassay (EIA), was evaluated for direct detection of antigens from Neisseria gonorrhoeae in clinical specimens. Results of the EIA were compared with those of culture and gram stain. Separate swab specimens for each procedure were collected from cervical secretions of 119 females and urethral exudate of 198 males (total, 324 specimens) seen in a sexually transmitted disease clinic. Cultures from 132 specimens were positive for gonococci (27 women and 105 men), and gram stains of 119 specimens were positive (14 women and 105 men). The EIA test was positive for 145 specimens (34 women and 111 men). The sensitivities of EIA for females and males were 88.5% and 99.0%, respectively, as compared with 29.6% and 98.1% for gram stain. Specificities of EIA were 89.2% and 92.5%, while those of gram stain were 93.5% and 86.0%, respectively. Overall accuracy of the EIA was 89.1% for females and 95.9% for males. In veiw of the speed and independence from viable organisms of the EIA test, the lack of good gram stain sensitivity for females, and the inherent drawbacks of culture, the Gonozyme test, if used in conjunction with culture, can detect a higher percentage of patients with genital gonorrhea than can culture alone. However, EIA in its present form cannot be used as the sole criterion for a laboratory diagnosis of gonococcal disease because of the ethical and medicolegal consequences of an EIA-positive, culture-negative result.(ABSTRACT TRUNCATED AT 250 WORDS)