We present a new methodology to estimate rates of energy acquisition, maintenance, reproductive investment and the onset of maturation (four-trait estimation) by fitting an energy allocation model to individual growth trajectories. The accuracy and precision of the method is evaluated on simulated growth trajectories. In the deterministic case, all life history parameters are well estimated with negligible bias over realistic parameter ranges. Adding environmental variability reduces precision, causes the maintenance and reproductive investment to be confounded with a negative error correlation, and tends, if strong, to result in an underestimation of the energy acquisition and maintenance and an overestimation of the age and size at the onset of maturation. Assuming a priori incorrect allometric scaling exponents also leads to a general but fairly predictable bias. To avoid confounding in applications we propose to assume a constant maintenance (three-trait estimation), which can be obtained by fitting reproductive investment simultaneously to size at age on population data. The results become qualitatively more robust but the improvement of the estimate of the onset of maturation is not significant. When applied to growth curves back-calculated from otoliths of female North Sea plaice Pleuronectes platessa, the four-trait and three-trait estimation produced estimates for the onset of maturation very similar to those obtained by direct observation. The correlations between life-history traits match expectations. We discuss the potential of the methodology in studies of the ecology and evolution of life history parameters in wild populations.