There are a variety of ways to measure religious beliefs. However, there is evidence that some commonly used measures of religiosity function poorly for atheist and spiritual individuals. In the ALSPAC there are a several measures of religiosity based upon pre-existing scales, such as the Duke University Religion Index (DUREL) and the Brief Multidimensional Measure of Religiousness/Spirituality (BMMRS). The aim of the current study is to examine the latent structure of the items taken from these scales, for use in ALSPAC, and to explore whether these items are measurement invariant across religious, atheist and spiritual individuals. This will provide an insight into the way religiosity is measured in ALSPAC and inform future scale construction using these items.