This paper is one of a series describing the performance and accuracy of map-making codes as assessed by the Planck CTP working group. We compare the performance of multiple codes written by different groups for making polarized maps from Planck-sized, all-sky cosmic microwave background (CMB) data. Three of the codes are based on destriping algorithm, whereas the other three are implementations of a maximum-likelihood algorithm. Previous papers in the series described simulations at 100 GHz (Poutanen et al. 2006) and 217 GHz (Ashdown et al. 2006). In this paper we make maps (temperature and polarisation) from the simulated one-year observations of four 30 GHz detectors of Planck Low Frequency Instrument (LFI). We used Planck Level S simulation pipeline to produce the observed time-ordered-data streams (TOD). Our previous studies considered polarisation observations for the CMB only. For this paper we increased the realism of the simulations and included polarized galactic foregrounds to our sky model. Our simulated TODs comprised of dipole, CMB, diffuse galactic emissions, extragalactic radio sources, and detector noise. The strong subpixel signal gradients arising from the foreground signals couple to the output map through the map-making and cause an error (signal error) in the maps. Destriping codes have smaller signal error than the maximum-likelihood codes. We examined a number of schemes to reduce this error. On the other hand, the maximum-likelihood map-making codes can produce maps with lower residual noise than destriping codes.