Autonomous in situ sensors are required to monitor high-frequency nutrient fluctuations in marine environments on a mass-scale. We present a submersible, dual-chemistry sensor that performs multiple colourimetric assays simultaneously on a fluid sample for multi-parameter in situ analysis. Based on a highly configurable architecture that has been successfully deployed for several multi-month periods, the sensor utilizes 10 solenoid valves, 4 syringes, 3 stepper motors, 2 LEDs, 4 photodiodes, and “inlaid” microfluidics to permit optical measurements of microliter fluid volumes. Fluid pathways are machined into a modular two-layer microfluidic lab-on-chip (LOC) fabricated from poly (methyl methacrylate) (PMMA) with two parallel inlaid optical cells of 10.4 mm and 25.4 mm path lengths (1.7 µl and 4 μl, respectively). Different LOC designs can be used to implement a wide variety of colorimetric assays. We demonstrate application of our dual-chemistry sensor towards simultaneous measurement of nitrate and dissolved orthophosphate: two nutrients fundamental to primary production. The performance of the dual-species nitrate and phosphate “NP Sensor” is characterized first in a controlled laboratory environment. Combined nutrient standards containing nitrate and phosphate concentrations ranging from 2.5 µM–100 µM NO3− and 0.25 µM–10 µM PO43− were analyzed, reporting detection limits of 97 nM NO3− and 15 nM PO43−. Calibrations were repeated under 3 fixed temperature conditions, T = 5°C, 10°C, 15°C, to determine the temperature-dependent sensitivity relations for both species needed to calculate concentrations during field deployments. Finally, an 8-day field deployment in Fish Hatchery Park, NS, Canada followed, acquiring a total of 592 nitrate and dissolved orthophosphate measurements. An on-board combined nutrient standard was measured periodically to assess the in situ accuracy of the sensor, with an average relative uncertainty of 15% across the deployment. Measured nitrate and dissolved orthophosphate levels in the river reached as high as 10 µM and 3.6 µM, respectively. Fast Fourier transform analysis suggests a strong out-of-phase relationship between measured phosphate and water level, with a shared frequency peak in both data agreeing within a 3.2% difference. This trend is due to conventional mixing at the river mouth to neighboring Bedford Basin. A spike in the measured nitrate to phosphate (N:P) ratio was also observed, synchronized to a precipitation event and indicative of runoff. The novel sensor will enable high-frequency dual-nutrient monitoring in many aquatic environments.