Over the past 25 years, there have been two opposing views of how taste information is represented in the activity of gustatory neurons. One view, the across-fiber pattern (AFP) theory, postulates that taste quality is represented by the pattern of activity across the afferent population. Stimuli with similar tastes produce similar patterns of activity. The other view is that activity in a few distinct neuron types codes taste quality in a "labeled-line" fashion. Neurons responding best to sucrose, for example, would represent "sweetness," and those responding best to NaCl would code "saltiness." Some of these neuron types appear to have a biological significance, such as the NaCl-best cells, which receive input about sodium stimuli exclusively from an amiloride-sensitive epithelial ion channel. However, the relatively broad tuning of these neurons makes it unlikely that they are capable of unambiguously coding information about taste quality. Rather, these neuron types play a critical role in establishing unique AFPs that distinguish among taste stimuli. The relative activity across these cell types represent taste quality, much like the patterns of activity across broadly tuned photoreceptors code information about stimulus wavelength.