This paper provides a descriptive analysis of segmental distributions in the Māori lexicon. Focussing on the strict-CV subset of the lexicon, we examine co-occurrence restrictions of consonantal onsets and vowel nuclei of adjacent syllables. For consonants, we find that sequences that share the same place of articulation are under-represented. This shows a similarity avoidance effect in Māori, reported for other languages (Frisch et al., 2004; McCarthy, 1986). When we correct for the presence of reduplicants in the data-set, this under-representation includes sequences of identical consonants. Sequences of identical vowels are overrepresented, even when reduplicated syllables are taken into account. The results show that gradient phonotactic processes are operating in Māori beyond the categorical restrictions on syllable shape.