Several techniques have been introduced in the last decades for the dehydration and release of O2 from biogenic silica (opal-A) for oxygen-isotope analysis. However, only one silica standard is universally available: a quartz standard (NBS28) distributed by the IAEA, Vienna. Hence, there is a need for biogenic silica working standards. This paper compares the existing methods of oxygen-isotope analyses of opal-A and aims to characterize additional possible working standards to calibrate the δ18O values of biogenic silica. For this purpose, an inter-laboratory comparison was organized. Six potential working standard materials were analysed repeatedly against NBS28 by eight participating laboratories using their specific analytical methods. The materials cover a wide range of δ18O values (+23 to +43‰) and include diatoms (marine, lacustrine), phytoliths and synthetically-produced hydrous silica. To characterize the proposed standards, chemical analyses and imaging by scanning electron microscopy (SEM) were also performed. Despite procedural differences at each laboratory, all methods are in reasonable agreement with a standard deviation (SD) for δ18O values between 0.3‰ and 0.9‰ (1σ). Based on the results, we propose four additional biogenic silica working standards (PS1772-8: 42.8‰; BFC: 29.0‰; MSG60: 37.0‰; G95-25-CL leaves: 36.6‰) for δ18O analyses, available on request through the relevant laboratories. © 2011 Elsevier Ltd.