Molecular breeding, a powerful technique to increase crop yield, tries to predict yield by crop growth models with genotype specific, environment-independent yield components and environmental indices as inputs. A fluorescence-trait-based approach is presented to approximate some costly and time-consuming measurements of yield components. Temporal monitoring of chlorophyll a fluorescence resulted in fluorescence traits with high heritability (0.60–0.82) that could act as proxies for model inputs. Medium-sized Pearson's correlations were calculated between fluorescence traits, light-use efficiency (LUE), and fruit related parameters up to 0.53. Multi-trait quantitative trait locus (QTL) analyses identified genomic regions of pepper (Capsicum annuum L.) involved in the phenotypic variation of the fluorescence traits. Fluorescence QTLs found on linkage groups P6, P7, and P11 corresponded to QTLs for number of fruits, partitioning into fruits, and LUE. Fluorescence parameters within 1 min of the fluorescence response curve can thus be useful to approximate yield component traits.