In the optical fiber detection configuration, linearity (r 240.995) was established up to an Fe(II) concentration of 120.0 nmol L1 whereas for window detection the upper value decreased to 20.00 nmol L1 . Calibration curves resulted from the triplicate analysis of 4 or 5 Fe(II) standard solutions with concentrations equally distributed along the linear working range (Fig. 4). The LOD of 1 nmol L1 for both detection strategies was calculated as the Fe(II) concentration equivalent to the analytical signal obtained by adding the blank signal to 3 times the corresponding standard deviation [23]. Sensitivity was strongly related to the detection strategy because the surface area monitored by the PMT was significantly greater using the window scheme (at least 30 times higher). For window detection, the slope of a typical calibration curve (52637286 arbitratry units nmol1 L, Fig. 4) was 355 times higher than the bifurcated optical fiber strategy (14.870.2 arbitratry units nmol1 L). Nevertheless, the background signal (calibration curve intercept value) was also higher for window detection (15,53173164) than for optical fiber detection (116714) though significantly lower than the gain in sensitivity (The values in parenthesis corresponded to the slope and intercept values and respective limits of the 95% confidence levels intervals.) In our final experimental conditions, we observed intra and inter-day slope variations below 10% of the values presented here. Relative standard deviation (RSD%) measured the precision of the determination. For both optical fiber and window detection strategies, RSD values were under 6% for triplicate measurements of each seawater standard solution. We also analyzed seawater samples (originally containing Fe(II) below our LOD of 1 nmol L1 ) spiked with different levels of Fe(II). In the optical fiber detection configuration, we found apparent recoveries [24] between 98.5% and 100.1%, for Fe (II) concentrations of 40.00 and 80.00 nmol L1 , respectively. When we used window detection, the apparent recovery ranged between 91.6% and 101.5%, for Fe (II) concentrations of 5.00 and 15.00 nmol L1 , respectively. Although a full validation of the analytical method is beyond the scope of the present work, these values highlight the potential of the two detection strategies for the quantification of Fe(II) in seawater at nanomolar levels using mSI-LOV methodology. In con