Phosphorus is an essential element for plants and animals, playing a fundamental role in the production of biochemical energy. Despite its relevance, phosphorus is not commonly determined by instrumental neutron activation analysis (INAA), because P does not emit gamma-rays in its decay. There are alternative methods for the determination of phosphorus by INAA, such as the use of beta counting or the measurement of bremsstrahlung originated from the high energy beta particle from P. Here the determination of phosphorus in plant materials by measuring the bremsstrahlung production was further investigated, to optimize an analytical protocol for minimizing interferences and overcoming the poor specificity. Eight certified reference materials of plant matrices with phosphorus ranging between 171 and 5,180 mg kg were irradiated at a thermal neutron flux of 9.5 × 10 cm s and measured with a HPGe detector at decay times varying from 7 to 60 days. Phosphorus solutions added to a certified reference material at three levels were used for calibration. Counts accumulated in the baseline at four different regions of the gamma-ray spectra were tested for the determination of phosphorus, with better results for the 100 keV region. The Compton scattering contribution in the selected range was discounted using an experimental peak-to-Compton factor and the net areas of all peaks in the spectra with energies higher than 218 keV, i.e. Compton edge above 100 keV. Amongst the interferences investigated, the production of P from sulfur, and the contribution of Compton scattering should be considered for producing good results. [ABSTRACT FROM AUTHOR]