This thesis investigates the modelling of massive protostellar envelopes in dust and molecular emission. A detailed study is undertaken for a sample of 30 targets believed to contain young high-mass protostars. SCUBA observations of the dust emission are used to help distinguish between the genuine high-mass protostars and clusters of lower mass protostars. Using the single power-law density structure, models are found that fit 16 sources, that are well-separated from other sources, in half of the 30 target regions. Indices of the power-law profiles are compared to values obtained from the literature. The distributions of the indices are similar but there are differences in the underlying distributions which show that observations at longer wavelengths give generally steeper profiles. Investigating the underlying nature of the sources between the fitted and non-fitted sources shows some interesting differences between the groups. These differences indicate that the modelled sources are more likely to be massive protostars while the other sources are probably lower mass protostars. Single power-law modelling is also used to investigate molecular emission in a second group of targets from the RMS survey. Dust models are obtained for six of the sources which show similar results to the sources described above. Molecular line emission from the targets is mapped and compared to the dust emission which shows they trace the same regions. The dust models are used as a basis to obtain the molecular abundances for the sources. Abundances from the modelling the lines are broadly similar values to those observed in other high-mass protostars and association with radio and maser emission indicates that high-mass star formation is definitely occurring in some of these regions.