Purpose: The variable deletion of word-final consonants is a well-known feature of African American English (AAE). This study aimed to show whether African American children exhibit an alveolar bias in their deletion of final voiceless stops as has been observed for their production of final nasals. Method: The data were extracted from more than 5,000 spontaneous utterances in the speech samples of 7 African American children at 32 to 36 months of age. Results: The final alveolar voiceless stop /-t/ was deleted significantly more often in word-final position than were /-p/ and /-k/ in both singleton and clustered contexts. The deletion of /-t/ in final clusters preceding other consonants at word boundaries contributed significantly to this bias. No significant differences were observed among the stops in their relative frequencies of deletion when a vowel sound followed or when the final stop was prepausal at word boundaries. Conclusion: African American children's deletion of final consonants is patterned even at an early age. It varies with whether the voiceless stop consonant is an alveolar sound or not and with the type of phonetic context in which the alveolar stop is embedded. This alveolar stop bias was attributed to phonetic and grammatic constraints on articulating final /-t/ relative to final /-p/ and /-k/. Clinical Implication: All final consonant deletion patterns should not be regarded as typical of African American children when assessing their speech even as early as age 3;0 (years;months).