Stellar mass, star formation rate (SFR) and gas metallicity are among the most important physical quantities that can be measured for a galaxy. Although the mass-metallicity and mass-SFR relations have been extensively studied, a tight relation between all three parameters has been discovered only recently. This 'fundamental metallicity relation' seems to be independent on redshift, suggesting the existence of some important equilibrium process that governs the evolution of galaxies. In order to study this relation at different redshifts, however, it is necessary to compare galaxy populations with similar properties. This is particularly challenging, since at high redshift we can observe only intrinsically bright galaxies, which have a SFR much higher than what is measured for local galaxies. One solution to this problem is to take advantage of the large magnification caused by strong gravitational lensing. I will present a new study of 10 low-mass, low-SFR galaxies in the redshift range 1.5 < z < 3 that are gravitationally lensed by foreground galaxy clusters. We obtained near-infared spectra using TripleSpec at the Palomar 200-inch telescope and we detected multiple rest-frame optical emission lines, which allowed us to measure star formation rate and gas metallicity. With this new data we can constrain the high-redshift fundamental metallicity relation at masses and star formation rates never probed before.