We tested the nanomaterial release from composites during two different mechanical treatment processes, automated drilling and manual sawing. Polyurethane (PU) polymer discs (1-cm thickness and 11-cm diameter) were created using different nanomaterial fillers: multiwall carbon nanotubes (MWCNT), carbon black (CB), silicon dioxide (SiO2), and an unfilled PU control. Drilling generated far more submicron range particles than sawing. In the drilling experiments, none of the tested nanofillers showed a significant influence on particle number concentrations or sizes, except for the PU/MWCNT samples, from which larger particles were released than from control samples. Higher drilling speed and larger drill bit size were associated with higher particle counts. Differences between composites were observed during sawing: PU/CB released higher number concentrations of micro-sized particles compared to reference samples. When sawing PU/SiO2 more nanoparticle agglomerates were observed. Furthermore, polymer fumes were released during sawing experiments, which was attributed to the process heat. For both drilling and sawing, the majority of the aerosolized particles were polymer matrix materials containing nanofillers (or protruding from their surface), as evidenced by electron microscopic analysis. Results suggest that: (i) processes associated with higher energy inputs are more likely to result in higher particle release in terms of number concentration; (ii) nanofillers may alter release processes; and (iii) other types of released particles, in particular polymer fumes from high-temperature processes, must also be considered in occupational exposure and risk assessments.