Topographic brain mapping of evoked potentials can be used to localize abnormalities of cortical function. We evaluated the effect of sleep fragmentation on brain function by measuring the visual P300 waveform using brain mapping. Eight normal subjects (Epworth Score ± SD: 5 ± 3) underwent tone-induced sleep fragmentation and undisturbed study nights in a randomized cross-over design. Study nights were followed by topographic brain mapping using a visual information processing test and concurrent event-related potentials. Experimental sleep fragmentation did not significantly increase objective daytime sleepiness or lower cognitive performance on a battery of cognitive function tests (all P ≥ 0.1). There were no significant topographical delays in P300 latencies with sleep fragmentation (all P > 0.15). However, at sites Fz, F4, T3, C3, Cz and C4 the P300 amplitudes were reduced significantly after sleep fragmentation (all P < 0.05). A reduction in P300 amplitude has previously been interpreted as a decrease in attention. These reductions in P300 amplitudes with sleep fragmentation in frontal, central and temporal brain areas suggest that sleep fragmentation may cause a broad decrease in attention. Sleep fragmentation did not delay P300 latencies in any brain area, and so does not explain the delay in P300 latencies reported in sleep apnoeics.