Infants perceive speech and acquire language amidst noisy and complex auditory environments. Thus, elucidation of the cognitive mechanisms governing speech perception under noisy conditions is crucial. Cortical encoding of the speech envelope has been one approach used to study speech-in-noise perception in adults. For infants, research shows that Infant Directed Speech (IDS) facilitates cortical encoding of the speech envelope in quiet conditions more than adult direct speech. However, it is unclear whether infants are able to track the IDS speech envelope amidst competing speech. To investigate this, we recorded the neural responses from 40 typically-hearing infants (20 seven-month-olds, 20 eleven-month-olds) to continuous IDS using electroencephalography (EEG) in three conditions: Quiet, Co-located Noise, and Separated Noise. The target stimuli consisted of naturally recorded IDS produced by two female English speakers. The noise stimuli consisted of a four-person babble constructed from audiobooks read by 2 male and 2 female English speakers. We presented stimuli at an overall level of 70 dB SPL via speakers placed at 0°, +90°, and -90° azimuth to infants sitting on a caregiver’s lap in a sound-attenuated booth. Our team analyzed EEG signals using the Multivariate Temporal Response Function (mTRF) toolbox in MATLAB. This backward modeling approach assesses whether the stimulus envelope can be reconstructed based on the recorded neural responses. Reconstruction accuracies greater than chance were observed in all three conditions for the majority of infants, suggesting that we were able to decode the speech envelope in both quiet and noise. Participants demonstrated the capacity to process speech, even amidst competing auditory stimuli, emphasizing speech perception competencies from an early developmental stage. These results support using the envelope model and mTRF method as a feasible method for investigating the development of speech-in-noise perception in infants and young children.