Triggers of indeterminate results from interferon-gamma release assays (IGRA) in patients with rheumatic diseases are still elusive. The aim of the present study was to describe predictors of indeterminate results from IGRA in the field of rheumatology. This cross-sectional study was retrospectively performed by using a database of patients with a request for QuantiFERON-TB Gold-In Tube test (QFT-GIT) for screening of latent tuberculosis infection. The study cohort included 631 patients with rheumatic diseases. All variables influencing indeterminate QFT-GIT results were investigated by logistic regression analysis. The overall frequency of indeterminate IGRA results was 6.8 % (43/631). Those with indeterminate results were more likely to be aged ≥70 years, female, visitors in winter, suffering from systemic lupus erythematosus (SLE), and using sulfasalazine or a tumor necrosis factor (TNF)-α inhibitor. In addition, a longer incubation time of >6 h increased the odds ratio of indeterminate IGRA results. In contrast, the automated ELISA processor, ankylosing spondylitis, and the use of a non-steroidal anti-inflammatory drug decreased the likelihood of indeterminate IGRA results. Lymphopenia, thrombocytopenia, anemia, and hypoalbuminemia were significantly associated with indeterminate IGRA results. Multivariate analysis revealed that SLE, use of sulfasalazine or a TNF-α inhibitor, and a manual ELISA system were significantly independent predictors of indeterminate IGRA results. The proportion of indeterminate results in patients with rheumatic diseases is not infrequent. Careful attention to the pre-analytical conditions should minimize the indeterminate results. Automation of the ELISA process seems to be a promising solution to decrease the rate of indeterminate response.