This article investigates how variations in aerosol loading affect droplet activation and the production of warm rain in shallow cumulus clouds. It is also examined whether the resulting differences can be identified and quantified using 35 GHz Ka-band radar signatures and how these relationships depend on the cloud life-cycle stage. Three idealized large-eddy simulations, T13 (clean, aerosol number concentration <em>N</em>ₐ = 100 cm−³), T53 (intermediate, <em>N</em>ₐ = 1000 cm−³), and T73 (polluted, <em>N</em>ₐ = 5000 cm−³), coupled to a Lagrangian cloud model (LES–LCM), are conducted with a fixed aerosol size-distribution shape under identical thermodynamic conditions based on the Barbados Oceanographic and Meteorological Experiment (BOMEX). Persistent particle tracking is used to reconstruct Lagrangian histories of super-droplets. Analysis of droplet activation and growth statistics shows that clusters of relative humidity (RH) values distinguish core-like and entrained-shell-like activation pathways. The simulation results further show that increased aerosol loading reduces the temporal window available for droplet activation, spectral broadening, and collision–coalescence. Accordingly, approximately 20.1 %, 15.3 %, and 15.0 % of trajectories reach drizzle size (<em>r</em> ≥ 40 µm) in the clean, intermediate, and polluted cases, respectively. Mixing diagnostics shift toward more deactivation-dominated behavior with increasing aerosol loading. These findings suggest that aerosol-dependent microphysical pathways remain detectable in Ka-band Doppler radar signatures. Therefore, Ka-band radar observations provide an observationally testable signature of delayed activation and suppressed warm-rain production, while it shall be noted that knowledge about the evolution state of the cloud system is essential for drawing conclusions about aerosol effects.