The layer-by-layer (L-b-L) deposition of oppositely charged polyelectrolytic macroinitiators has been demonstrated on planar silica substrates. The build-up of the macroinitiator multilayers was monitored by ellipsometry (up to 21 layers) and dual polarization interferometry (up to 17 layers) and good agreement was found between these techniques. The increase in L-b-L thickness was approximately linear, with an average thickness of 2.3 angstrom per layer of deposited macroinitiator. Surface-initiated ATRP of a model nonionic methacrylic monomer, 2-hydroxyethyl methacrylate (HEMA) in a 1: 1 methanol/water mixture was conducted at ambient temperature. Increasing the number of macroinitiator layers led to a significant increase in PHEMA brush thickness up to 110 nm, which is attributed to the greater surface grafting density. PHEMA brush thicknesses obtained after 22 h showed a linear dependence on the number of layers of deposited macro-initiator, with all layers exhibiting near-identical growth kinetics. X-ray photoelectron spectroscopy was used to monitor L-b-L assembly and also to confirm PHEMA growth. This technique indicated the loss of small counterions from the multilayers during L-b-L deposition and confirmed an increase in the surface density of bromoester initiator groups as the number of deposited macroinitiator layers was increased. For 17 macroinitiator layers, the bromoester initiator density is estimated to be similar to 4.9 +/- 0.2 nm(-2) from the DPI data. This is comparable to that calculated for ATRP initiator monolayers obtained by either thiol or silane chemistry. Ellipsometry suggested that the macroinitiator multilayers were weakly hydrated prior to the in situ HEMA polymerization. AFM studies indicated that the PHEMA brushes had appreciable surface roughness, but this roughness became negligible compared to the brush thickness with increasing macroinitiator layers.