Abstract Background Instruments to examine patients’ communication skills are impractical for daily clinical practice in primary health care, and it is assumed that health care professionals are always aware of the communication and comprehension status of patients. Objective To design and examine the psychometric properties of a new questionnaire (AsCkS) to screen patients’ risk of low communication and comprehension skills. Methods Designing the new questionnaire involved: (i) a content validity study to generate a set of items, (ii) administration of a questionnaire in a sample of patients and (iii) study of the psychometric properties using a Rasch probabilistic model. The test probability function was used to detect patients at risk of having low communication and comprehension skills, and the associated factors were studied using multivariate logistic regression. Results Five items were generated and 369 patients screened. In the Rasch analysis, one item was removed due to a lack of goodness-of-fit (Outfit = 3.64). The final set of four items showed good reliability (person separation index = 0.90). Convergent validity was moderate for the health literacy construct (r = 0.53) and low for the patient activation (r = 0.38): 77 patients (20.86%) were at risk of having low communication and comprehension skills. After adjustment, the associated factors were age (p < 0.001) and lower educational level (p < 0.001). Conclusion The AsCkS is a reliable and valid instrument and may be used to detect patients requiring a greater effort by the physician to communicate health care messages.