There is considerable interest in assessing the magnitude of carbon sources and sinks for agricultural lands, grasslands, and forests. In this paper, we propose a novel technique to remotely assess CO2 fluxes in maize using reflectances (ρ) in two spectral channels either in the green around 550 nm or in the red edge near 700 nm and the NIR (beyond 750 nm). Differences of reciprocal reflectances <(ρGreen)-1 - (ρNIR)-1> and <(ρRedEdge)-1 - (ρNIR)-1> accounted for more than 90 percent of the variability in mid-day canopy photosynthesis of irrigated maize. The technique was validated by an independent data set; root mean square error in predicting mid-day canopy photosynthesis by <(ρRedEdge)-1 - (ρNIR)-1> was 0.17 mg/m2/s and 0.2 mg/m2/s by <(ρGreen)-1 - (ρNIR)-1>.