Over the past 3.5 yr, we have examined 195 head and neck fine-needle aspiration (FNA) specimens from three diverse medical settings. Specimens were collected in saline solution or Saccomanno fixative and processed using cytocentrifugation or membrane filtration. This allowed us simultaneously to perform cytologic evaluations, special stains, and immunologic marker studies from a single specimen. Good correlation between clinical experience with FNA and obtaining satisfactory specimens was demonstrated. Our sensitivity (89%) and specificity (94%) reflect problems associated with specimen collection in a training environment where clinician experience with the procedure is low. A definitive diagnosis was possible in most cases, and the treatment plan was often based on the FNA results. In patients without a history of a primary malignancy, a FNA done early in the clinical course helped direct the initial workup of the patient, saving time and expense.