The evidence in adults suggests that at a cortical level simple and complex sounds are processed by partly divergent subsystems. In children, central processing of sounds differing in complexity has not been investigated. Therefore, the present study examined preconscious discrimination of the differences in sound frequency and duration as a function of sound complexity in 8-10-year-old children. A mismatch negativity (MMN) component of auditory event-related potentials was elicited in a paradigm where 'deviant' (rare) stimuli were either shorter in duration or higher in frequency than the 'standard' (repetitive) sounds. Vowels and vowel-matched complex and simple tones were presented in separate sequences. The stimulus complexity effects were sizable and appeared as larger areas and shorter and more consistent latencies of the MMNs, elicited by more complex stimuli. In addition, the vowel frequency MMN showed left hemisphere preponderance compared to the complex tone frequency MMN. No such effect was found for the duration decrement MMNs. In addition, the complex tone duration decrement MMN was distributed posteriorly to either the vowel or sinusoidal tone MMNs. A late discriminative negativity, LDN, did not show consistent effects of sound complexity. In conclusion, acoustically rich sound content facilitates auditory sensory discrimination in 8-10-year-old children. The sound 'speechness' effects were not as robust though present. Unlike adults, children demonstrated high intersubject variability in discriminating spectrally poor, but not rich, sounds. The discrimination of the sound duration appears to differ from that of the sound frequency in nature and, consequently, in the neural substrates.