Genetic and environmental differences are by far the most studied drivers underlying phenotypic variation. However, a growing number of studies finds among individual variation that is unexplained by genes or environment. Up to now, it remains an open question whether such seemingly stochastic variation has fitness consequences. To address this question, we performed a tightly controlled long term life history experiment with naturally clonal fish (Poecilia formosa) separated directly after birth into identical environments. Maintaining highly standardized conditions for 280 days, we first recorded individuals for 10 hours per day over the first 28 days of their lives, characterizing entire early life behavioural profiles. We then measured the complete reproductive profiles of these individuals over an average of 4.5 successive broods per individual, quantifying in total 2522 offspring from 152 broods. We find that (i) individuals differ consistently in the size of offspring and broods produced over consecutive broods, despite being genetically identical and being raised in identical environments. (ii) These differences are observed even when controlling for trade-offs between brood size, offspring size and reproductive onset, indicating that individuals differ in life history productivity. (iii) We also find strong early-life behavioural individuality in both activity and feeding pattern, with among-individual differences in feeding being predictive of growth, and consequently reproduction: individuals that consistently spend more time feeding become larger and larger individuals consistently produce larger offspring. Our findings provide experimental evidence that processes unexplained by genetic and environmental differences translate into predictable differences in life history measures and ultimately reproductive fitness.