Time-resolved laser-induced incandescence is used to infer the size distribution of gas-borne nanoparticles from time-resolved pyrometric measurements of the particle temperature after pulsed laser heating. The method is highly sensitive to aspects of the measurement strategy that are often not considered by practitioners, which often lead to discrepancies between measurements carried out under nominally identical conditions. This paper therefore presents a well-documented calibration procedure for LII systems and quantifies the uncertainty in pyrometric temperatures introduced by this procedure. Calibration steps include corrections for: (1) signal baseline, (2) variable transmission through optical components, and (3) detector characteristics (i.e., gain and spectral sensitivity). Candidate light sources are assessed for their suitability as a calibration reference and the uncertainty in calculated calibration factors is determined. The error analysis is demonstrated using LII measurements made on a sooting laminar diffusion flame. We present results for temperature traces of laser-heated particles determined using two- and multi-color detection techniques and discuss the temperature differences for various combinations of spectral detection channels. We also summarize measurement artifacts that could bias the LII signal processing and present strategies for error identification and prevention.