Boundary tones are pitch patterns conveying pragmatic and paralinguistic information. In Putonghua (PTH), they are the results of the interaction between tone and intonation and can be categorized into two types: simultaneous addition boundary tone (SiABT) and successive addition boundary tone (SuABT). PTH differentiates word meanings using tonal contrasts, challenging learners whose native languages lack such contrasts. This study examines whether PTH-learning preschoolers (PTH learners) from south Xinjiang Uyghur Autonomous region can correctly decode the lexical tone categories intertwined with SiABT and SuAB. The correct decoding must be based on the integration of segmental, tonal, and semantic information. Nineteen children aged 3;9–5;3 with about 7 mouths of PTH learning in kindergarten, participated in a word recognition task with two conditions. In Condition 1, 24 familiar monosyllabic words in 12 minimal pairs contrasting in lexical tones and conditioned with SiABT or falling SuABT were used. The results indicated that for the PTH learners, the SuABT facilitated lexical tone decoding, especially for boys. However, compared to younger PTH-speaking children (1;6–2;8), the PTH learners had weaker tonal category decoding ability. In Condition 2, 8 familiar monosyllabic words in 4 pairs without phonological contrast were used. The accuracy rates for PTH learners were similar to those of PTH speakers, except that the SuABT didn’t improve the accuracy rate for PTH learners, which indicated that the learners may depend primarily on segmental information and face challenges in integrating the tonal and pragmatic functions of intonation. The results support the automatic selective perception (ASP) model for L2 learners and suggest that tone acquisition is challenging for PTH learners, offering an effective method for teaching tones to preschool learners in the Xinjiang Uyghur Autonomous Region.