We present a comprehensive theoretical study, using a semi-analytical model within the standard LCDM framework, of the photometric properties of the progenitors of present-day early-type galaxies in the redshift range 0 < z < 1. We explore progenitors of all morphologies and study their characteristics as a function of the luminosity and local environment of the early-type remnant at z = 0. In agreement with previous studies, we find that, while larger early-types are generally assembled later, their luminosity-weighted stellar ages are typically older. In dense cluster-like environments, ∼70 percent of early-type systems are "in place" by z = 1 and evolve without interactions thereafter, while in the field the corresponding value is ∼30 percent. Averaging across all environments at z ∼ 1, less than 50 percent of the stellar mass which ends up in early-types today is actually in early-type progenitors at this redshift, in agreement with recent observational work. The corresponding value is ∼65 percent in clusters, due to faster morphological evolution in such dense environments. We develop probabilistic prescriptions which provide a means of including spiral (i.e. non early-type) progenitors at intermediate and high redshifts, based on their luminosity and optical colours. For example, we find that, at intermediate redshifts (z ∼ 0.5), large (M V < −21.5), red (B − V > 0.7) spirals have ∼75-95 percent chance of being an early-type progenitor, while the corresponding probability for large blue spirals (M B < −21.5, B − V < 0.7) is ∼50-75 percent. The prescriptions developed here can be used to address, from the perspective of the standard model, the issue of "progenitor bias", whereby the exclusion of late-type progenitors in observational studies can lead to inaccurate conclusions regarding the evolution of the early-type population over cosmic time. Finally, we explore the correspondence between the true "progenitor set" of the present-day early-type population -defined as the set of all galaxies that are progenitors of present-day early-types regardless of their morphologies -and the frequently used "red-sequence", defined as the set of galaxies within the part of the colour-magnitude space which is dominated by early-type objects. We find that, while more massive members (M V ≤ −21) of the "red sequence" trace the progenitor set reasonably well, the relationship breaks down at fainter luminosities (M V ≥ −21). Thus, while the results of recent observational studies which exploit the red sequence are valid (since they are largely restricted to massive galaxies), more care should be taken when deeper observations (which will probe fainter luminosities) become available in the future.