Fluorescence spectroscopy is a promising technology for detection of epithelial precancers and cancers. In preparation for a multicenter phase II screening trial, a pilot trial was conducted to test data collection and patient examination procedures, use data forms, time procedures, and identify problems with preliminary data analysis. Women 18 years of age and older underwent a questionnaire, a complete history, and a physical examination, including a pan-colposcopy of the lower genital tract. A fiber-optic probe measured fluorescence excitation-emission matrices at 1-3 cervical sites for 58 women. The data collection procedures, data forms, and procedure times worked well, although collection times for all the clinical data take an average of 28 min. The clinical team followed procedures well, and the data could be retrieved from the database at all sites. The multivariate analysis algorithm correctly identified squamous normal tissue 99% of the time and columnar normal tissue only 7%. The assessment of ploidy from monolayer samples was not accurate in this small sample. The study was successful as a pilot trial. We learned who participated, who withdrew, how often abnormalities were present, and that algorithms that have worked extremely well in previous studies do not work as well when a few study parameters are changed. The current algorithm for diagnosis identified squamous normal tissue very accurately and did less well for columnar normal tissue. Inflammation may be an explanation for this phenomenon. Fluorescence spectroscopy is a promising technology for the detection of epithelial precancers and cancers. The screening trial of fluorescence and reflectance spectroscopy was successful.