Not exactly hot off the press, but here's an interesting review article on neuroscientific insights into intelligence (Gray & Thompson, 2004, Nat Rev Neurosci, 5, 1-13). The authors focus on fluid intelligence (Gf), which is moderately correlated with brain volume. Neuroimaging and neuropsychological research implicate the lateral prefrontal cortex (PFC) as a brain region that is active during Gf tests and related to individual differences in Gf. The review explains that grey matter volume in the lateral PFC is both correlated with fluid intelligence and largely under genetic control, although it should be remembered that environmental input has been related to differences in brain morphology. Individual genes associated with Gf are difficult to pin down, while environmental factors are rather more easily detected (e.g., lead exposure).
On the vexed question of possible group-based differences, the authors propose that it is better to buttress ethical standards than to censor research in this field entirely. For example, they propose that data should only be retained if participants give their active support to a study's aims after a thorough debriefing.