Objective About 10 years ago, Gratz and Roemer (2004) introduced the Difficulties in Emotion Regulation Scale (DERS), a 36‐item self‐report instrument measuring 6 areas of emotion regulation problems. Recently, Bjureberg et al. (2015) have introduced a new, briefer version of the DERS comprising only 16 of the 36 items included in the original version. Because no studies have yet cross‐validated the recently introduced 16‐item DERS and the 36‐item DERS has never been tested in Brazil, we sought to inspect the psychometric properties of scores from both DERS versions with a nonclinical Brazilian sample. Method Participants were 725 adult volunteers aged 18–70 years (mean = 30.54, standard deviation = 10.59), 82.3% of whom were women. All were administered the DERS along with a number of other self‐report and performance‐based instruments. Data analyses inspected internal consistency, factor structure, and convergent as well as divergent validity of scores from both DERS versions. Results Results show that scores from both DERS versions possess good psychometric properties. Interestingly, both versions correlated, in the expected direction, with psychopathology and showed no significant correlations with cognitive measures. Like in other studies, however, the Awareness factor of the 36‐item DERS did not produce optimal validity and reliability indexes. Conclusion Taken together, our findings indicate that the 16‐item DERS may be preferred over the 36‐item version and provide additional support to the differentiation between emotion regulation and cognitive tasks of emotional perception and abstract and verbal reasoning.