Automated count of Nile Red fluorescent microplastics allows fast and reliable quantification. However, factors involving staining, digital camera conditions and settings introduce variability to the results. The objective of this paper is to identify and propose solutions to these factors and improve on the previous MP-VAT script. While removal of digital sensor defects had little influence on results and staining can be reduced to 5 min, Nile Red concentrations cannot be reduced <0.01 mg mL−1, the 470 nm LED lantern emission must be >1600 lx, and photographic conditions should be maintained as stable as possible ideally improving the filter membrane area and using the recommended settings of 2 s, ISO100, F5.6. It was also found that Nile Red can be removed from microplastics using acetone or hydrogen peroxide with iron. More importantly, both particles and fluorescent are lost with time and thus quantification should be conducted within a week. Finally, MP-VAT 2.0 was developed to remove unselected areas and to identify only red particles, excluding white reflections from quantification. This updated version of MP-VAT produced improved recovery rates of 98.2 ± 6.9 for spiked samples and 95.9 ± 10.3 on actual environmental samples, presenting a cheap and reliable complementary method for microplastic identification. Display omitted •Stable conditions are needed for proper quantification of fluorescent microplastics.•Staining with 0.01 mg mL−1 of Nile Red under 5 min produced good fluorescence.•Camera should be set to 2 s, ISO100, F5.6 and focused to maximize sample area.•Light emission from the 470 nm LED lantern should be >1600 lx.•Improved MP-VAT 2.0 presents recovery rates of 95.9 ± 10 in environmental samples.