Based on laboratory feeding experiments, this effort focuses on demonstrating the feasibility of using molecular markers coupled with statistical modeling for estimation of starvation-induced mortality in larval Atlantic cod. Our earlier studies have established that bulk RNA, 18S ribosomal RNA (rRNA) and metallothionein (MT) mRNA levels may serve as specific markers for starvation in cod larvae. A series of laboratory experiments will be completed to further characterize the responses of 18s rRNA and MT to changes in feeding, development and growth in Atlantic cod. Laboratory results (molecular biology and survival data) will be used to calibrate a Cox regression model for estimation of starvation mortality risks in larval Atlantic cod. The nonproportional hazards Cox regression model chosen for this study is commonly used by the life insurance industry for establishing risks and premium rates. It can accommodate time-dependent covariates, time-lagged effect, and a stepwise change in hazard rates over the lifespan of the organism studied. Our earlier work established that significant decreases18s rRNA levels can be used as a marker for the initial stages of starvation in Atlantic cod larvae. Dramatic increases in MT mRNA levels also were shown to be indicative of starvation. Both markers will be used to pinpoint recovery from starvation after refeeding. The power of mortality risks estimation will greatly increase if recovery from starvation can be established. A single marker alone may not be able to define the `point-of-no-return` in starvation with sufficient confidence. Animal and human studies have shown that dietary deprivation leads to polysome breakdown and a reduction in the number and activity of ribosomes. Refeeding after starvation results in significant increase in the levels of rRNAs. However, when starvation has exceeded the `point-of-no-return`, i.e. rRNA levels fail to rebound after refeeding, mortality sets in. To the extent that Atlantic cod is a suitable model for other fish and invertebrates that produce planktonic larvae, this effort will provide a picture of the molecular response of larvae to changing feeding conditions and provide tools to determine growth, starvation mortality and feeding conditions. These studies should provide insight into the effects of environmental variability on production of fish.