Early childhood education institutions in remote areas such as West Kutai face challenges in identifying children's learning styles due to limited access to professional psychologists. This web-based psychoeducational system was developed to detect early childhood learning styles using the Fuzzy Tsukamoto method, which is capable of producing clear outputs from linguistic inputs. The system is built with a three-tier architecture and uses 20 Likert-scale questionnaire items to identify three learning style dimensions: Visual, Auditory, and Kinesthetic. The Fuzzy Tsukamoto algorithm is implemented through four stages: score calculation, fuzzification using trapezoidal functions, inference based on 27 rule bases, and defuzzification with weighted average. The system features an interactive questionnaire, result visualization, child development monitoring, and an administrator dashboard. Test results show algorithm accuracy of 99.98% based on validation with psychologist diagnoses, and user satisfaction of 4.8/5.0 from User Acceptance Testing with early childhood teachers. This system provides a practical and affordable solution for early childhood institutions to identify children's learning styles and design more personalized learning strategies without relying on expensive external consultations.