This paper provides a more detailed description of normal tongue palate contact patterns for the occlusion phase of velar stops than currently exists. The study used electropalatography (EPG) to record seven normally speaking adults' contact patterns of voiceless velar stops in nine VkV contexts. A variety of EPG indices measured: per cent complete closures across the palate; place of articulation; articulatory distance between /k/ and /t/; and amount of contact. Complete EPG closure occurred in the majority (81%) of tokens. Some speakers almost always had complete closures regardless of context, whereas other speakers produced them only with high front vowels. As expected, place of articulation and amount of contact were heavily influenced by vowel context. The most fronted and most contact occurred in /i/ contexts and the most retracted and least contact in /a/ contexts. There was considerable interspeaker variability on all measures, particularly in the precise location of velar placement and articulatory distance between alveolar and velar placement. The implications for diagnosing and treating abnormal velar articulations are discussed. (Contains 6 figures.)