We described the background and the development of a new measure of existential loneliness, the Existential Loneliness Questionnaire (ELQ). Specifically, we analyzed the items of the preliminary version of the ELQ (ELQ-P) using methods based on item response theory (the Rasch model) and examined the convergent and discriminative validity of the ELQ in a sample of 47 HIV-infected women. Item analysis produced an ELQ version consisting of 22 items that were internally consistent and performed well in measuring an underlying construct conceptualized as existential loneliness. In addition, the ELQ discriminated well between symptomatic and asymptomatic HIV-infected women. The ELQ correlated strongly with measures of depression, loneliness not identified as existential and purpose-in-life and moderately strongly with a measure of hopelessness. Holding constant depression scores, the correlation between the ELQ and loneliness not identified as existential was significantly attenuated. Limitations of the study include the small sample size, which precluded an analysis of the dimensional structure of the ELQ. © 2002 Wiley Periodicals, Inc. J Clin Psychol 58: 1183–1193, 2002.