The early development of vocalic and consonantal production in Mandarin-learning infants was studied at the transition from babbling to producing first words. Spontaneous vocalizations were recorded for 24 infants grouped by age: G1 (0 ; 7 to 1 ; 0) and G2 (1 ; 1 to 1 ; 6). Additionally, the infant-directed speech of 24 caregivers was recorded during natural infant-adult interactions to infer language-specific effects. Data were phonetically transcribed according to broad categories of vowels and consonants. Vocalic development, in comparison with reports for children of other linguistic environments, exhibited two universal patterns: the prominence of [epsilon] and [schwa], and the predominance of low and mid vowels over high vowels. Language-specific patterns were also found, e.g. the early appearance and acquisition of low vowels [a]. Vowel production was similar in G1 and G2, and a continuum of developmental changes brought infants' vocalization closer to the adult model. Consonantal development showed two universal patterns: labials and alveolars occurred more frequently than velars; and nasals developed earlier than fricatives, affricates and liquids. We also found two language-specific patterns: alveolars were more prominent than labials and affricates developed early. Universal and language-specific characteristics in G1 continued to be prominent in G2. These data indicate that infants are sensitive to the ambient language at an early age, and this sensitivity influences the nature of their vocalizations.