Two Greek Pleistocene tufa stromatolites were examined petrographically and with stable isotope geochemistry to determine whether calcite spar is of primary or diagenetic origin. The younger (ca 100 ka) tufa from Zemeno primarily is micritic, with primary columnar calcite spar restricted to areas immediately above chironomid larval tubes. This relationship suggests that chironomid larval feeding behaviour is responsible for the development of Zemeno tufa columnar calcite, probably involving biological substances smeared onto the tufa surface. Most micritic crystals are not suitably oriented to allow later post-depositional growth resulting in columnar fabrics. The older (ca 1 Ma) predominantly sparry tufa from Nemea contains some chironomid tubes and organic cyanobacterial filaments preserved in crystal fans but also contains many fabrics found in primary speleothem spar. Columnar spar here is unlikely to be the result of post-depositional crystal growth. A comparison of stable isotopic trends between the two tufa deposits suggests that both contain interpretable seasonal trends and implies little or minor post-depositional alteration of either tufa. Consequently, there is no basis for the common assumption that sparry tufa fabrics must be of diagenetic origin.