Evolution of ageing requires mutations with late-life deleterious effects. Classic theories assume these mutations either have neutral (Mutation Accumulation) or beneficial (Antagonistic Pleiotropy) effects early in life, but it is also possible that they start out as mildly harmful and gradually become more deleterious with age. Despite a wealth of studies on the genetics of ageing, we still have a poor understanding of how common mutations with age-specific effects are and what ageing theory they support. To advance our knowledge on this topic we measure a set of genomic deletions for their heterozygous effects on juvenile performance, fecundity at three ages, and adult survival. Most deletions have age-specific effects, and these are commonly harmful late in life. Many of the deletions assayed here would thus contribute to ageing if present in a population. Taking only age-specific fecundity into account, some deletions support Antagonistic Pleiotropy, but the majority of them better fit a scenario where their negative effects on fecundity become progressively worse with age. Most deletions have a negative effect on juvenile performance, a fact which strengthens the conclusion that deletions primarily contribute to ageing through negative effects that amplify with age.