The Star Formation History of the Universe changes with time and has a clear maximum around redshift z ~ 2, but it is not clear what drives the subsequent decline. Recent studies have suggested a possible link between the cosmic density of H2 - the most abundant molecule in the Universe - and the Star Formation History of the Universe (Decarli et al. 2019, Keating et al. 2016, Riechers et al. 2018). The second most abundant molecule, still linked to star formation, is CO and its rotational transitions are bright and relatively easy to observe with ALMA. By studying the CO emission in galaxies we can quantify the molecular gas density in galaxies. In my talk I will present a blind CO emission-line survey using ALMACAL - a extensive science project utilizing ALMA calibration data for scientific purposes. ALMA calibrators are usually bright sub-mm quasars (mostly blazars) and are relatively straightforward to subtract from the data, leaving an image or data cube of the calibrator field. Using the ALMACAL cubes with the longest integration time, we build a cosmological significant volume over which we can blindly search for CO emission lines. Thanks to a uniqueness of the ALMACAL dataset we are able to study galaxies over a wide area, and are not sensitive to the effects of cosmic variance.