Rooftop photovoltaic (PV) systems modify urban surface properties and can influence roof temperatures, heat fluxes, and both outdoor and indoor air temperatures. Yet, reported impacts of PV on the urban atmosphere remain inconsistent, highlighting the need for a more systematic modeling approach. In this paper, we present a sophisticated energy budget parameterization of building-applied PV in the microscale large-eddy simulation (LES) model PALM. It accounts for the PV-roof air gap, capturing thermal and radiative interactions, ventilation, and PV material properties. Validation against a five-month measurement campaign in offline energy-balance mode (no atmosphere) shows high accuracy ( R ² = 0.93), with 80 % of simulated PV temperatures falling within ± 4 K of observations. In a first building-resolving LES application for a dense urban area (LCZ 02), we found that area-wide PV deployment reduces near-surface air temperatures (24-h averages) compared to black roofs (+0.74 K), gravel roofs (+0.45 K), and terracotta roofs (+0.32 K). PV and red roofs show negligible differences (−0.01K). Nighttime effects are more complex: PV roofs cool faster in the early evening but may slightly increase temperatures around dawn due to low ventilation and radiation trapping in PV-roof air gap. White roofs remain constantly cooler than PV (−1.09K). This study represents the first integration of rooftop PV into microscale LES simulations resolving individual buildings within a 2 km × 2 km domain, emphasizing the role of detailed energy balance and microscale flows in assessing PV impacts and guiding urban energy planning.