This study investigated Mandarin-speaking children’s spontaneous speech production of contrastive focus in five-syllable SVO declaratives (constructed as Adjective1+Noun1+Verb+Adjective2+Noun2). Ten children aged 4;7-6;0 participated in a picture-guessing game with their parents, where they were prompted to produce the target structure with contrastive focus at different positions. The results showed a generally accurate placement of the focal accent, which improved with age. Nevertheless, 10.2% of the total contrastive foci were misplaced in wrong prosodic words, and 8.3% in correct prosodic words but on a wrong syllable. Additionally, 37.5% of contrastive foci on Noun2 were realized at incorrect positions, while very few on Verb were misplaced. Misaccenting Noun2 was the most common error, followed by Adjective1 and Adjective2. Acoustic analysis revealed children’s awareness of using prosodic features including pitch range expansion, durational adjustment, and post-focus compression to encode contrastive focus. However, the performance varied across individuals and developed gradually. Furthermore, hesitation pauses were found in a third of the utterances. The findings suggest that the adult-like mastery of contrastive focus realization is still developing by age 6, which is closely related the complex interaction of lexical and phrase-level prosody in Mandarin Chinese.