Extensive research shows temperature to be the primary environmental factor controlling the phyllochron, or rate of leaf appearance, of wheat (Triticum aestivum L.). Experimental results suggest that soil temperature at crown depth, rather than air temperature above the canopy, would better predict wheat leaf appearance rates. To test this hypothesis, leaf appearance in spring wheat ('Nordic') was measured in a 2-year field experiment (Nunn clay loam soil; fine, smectitic, mesic Aridic, Argiustoll) with three planting dates and two soil temperature treatments. One temperature treatment (denoted +3C) consisted of heating the soil at crown depth to 3 degrees C above the ambient soil temperature (denoted +0C). Main stem cumulative leaf number was measured at least weekly until flag leaf emergence. Leaf appearance was essentially linear with both air and soil growing degree-days (GDD), although there was a stronger linear relationship with soil GDD in the +0C plants than in +3C plants. A weak positive relationship between planting date and the phyllochron was observed. Unexpectedly, we found that heating the soil did not increase the rate of leaf appearance, as the paradigm would predict. To explain these results, we propose extending the paradigm in two ways. First, three processes are involved in leaf appearance: (1) cell division at the shoot apex forms the primordium; (2) cell division in the intercalary meristem forms the cells that then (3) expand to produce the leaf. Cell division is predominantly controlled by temperature, but cell expansion is considerably more affected by factors other than temperature, explaining the influence of other factors on the phyllochron. Secondly, the vertical distribution of the two meristems and region of cell expansion occur over a significant distance, where temperature varies considerably, and temperature at a specific point (e.g. crown depth) does not account for the entire temperature regime under which leaves are developing.