ABSTRACT Objectives: The Bristol Stool Form Scale (BSFS) is inadequate for non‐toilet trained children. The Brussels Infant and Toddler Stool Scale (BITSS) was developed, consisting of 7 photographs of diapers containing stools of infants and toddlers. We aimed to evaluate interobserver reliability of stool consistency assessment among parents, nurses, and medical doctors (MDs) using the BITSS. Methods: In this multicenter cross‐sectional study (2016–2017), BITSS photographs were rated according to the BSFS. The reliability of the BITSS was evaluated using the overall proportion of perfect agreement and the linearly weighted κ statistic. Results: A total of 2462 observers participated: 1181 parents (48.0%), 624 nurses (25.3%), and 657 MDs (26.7%). The best‐performing BITSS photographs corresponded with BSFS type 7 (87.5%) and type 4 (87.6%), followed by the BITSS photographs representing BSFS type 6 (75.0%), BSFS type 5 (68.0%), BSFS type 1 (64.8%), and BSFS type 3 (64.6%). The weakest performing BITSS photograph corresponded with BSFS type 2 (49.7%). The overall weighted κ‐value was 0.72 (95% CI 0.59–0.85; good agreement). Based on these results, photographs were categorized per stool group as hard (BSFS type 1–3), formed (BSFS type 4), loose (BSFS types 5 and 6), or watery (BSFS type 7) stools. According to this new categorization system, correct allocation for each photograph ranged from 83 to 96% (average: 90%). The overall proportion of correct allocations was 72.8%. Conclusions: BITSS showed good agreement with BSFS. Using the newly categorized BITSS photographs, the BITSS is reliable for the assessment of stools of non‐toilet trained children in clinical practice and research. A multilanguage translated version of the BITSS can be downloaded at https://bitss‐stoolscale.com/ .