Background: Remote photoplethysmography (RPPG) is a technique in which we measure sub-cutaneous variations in blood flow, usually through a camera, to obtain physiological signals. Studies involving RPPG have increased in the past few years due to its numerous applications including remote healthcare, anti-spoofing, among others. While there have been many studies on how to increase RPPG’s accuracy of biomarkers predictions in a variety of settings, most of them are done using workstations, yet some of the most promising applications of RPPG probably would be on limited resources, low-power embedded systems. Aims: Understand effects of one of the most important design parameters in RPPG systems, sliding window (SW) size, in order to quantify the trade-off between computational cost in time and accuracy in root-mean-squared-error (RMSE). Methodology: We compared the results among 14 subjects from the UBFC-RPPG database of two face detection algorithms (Haar Cascade Classifier and Deep Learning-based Single-Shot Detector), two different ROI-selection schemes (skin-pixel segmentation and rectangular forehead), and four different RPPG processing algorithms (G [Verkruysse, 2008], PCA [Lewandoska, 2011], CHROM [de Haan, 2013] and POS [Wang, 2017]) across 20 SW sizes (1-20 s, step: 1 s). Results: For most configurations, RMSE decreases linearly as SW increases, until a certain point, while computational cost increases in an exponential way. Based on these results, we came up with a new and simple metric that takes into account both, as a means to design systems which make the best out of the available resources. We coined this metric as RMSE_lt=RMSE+alpha*log(t), a modification of RMSE which adds a weighted term for the log of the computation time. With correct tuning, we can choose alpha to reduce computational costs by 47%. Therefore, this metric could help to choose optimal SW size so as to obtain a better trade-off between resource use and accuracy.