Bateman's principle predicts the intensity of sexual selection depends on rates of increase of fecundity with mating success for each sex (Bateman slopes). The sex with the steeper increase (usually males) is under more intense sexual selection and is expected to compete for access to the sex under less intense sexual selection (usually females). Under Bateman and modern refinements of his ideas, differences in parental investment are key to defining Bateman slopes and thus sex roles. Other theories predict sex differences in mating investment, or any expenditures that reduce male potential reproductive rate, can also control sex roles. We focus on sexual behaviour in systems where males have low paternal investment but frequently mate only once in their lifetimes, after which they are often killed by the female. Mating effort (=terminal investment) is high for these males, and many forms of investment theory might predict sex role reversal. We find no qualitative evidence for sex role reversal in a sample of spiders that show this extreme male investment pattern. We also present new data for terminally-investing redback spiders (Latrodectus hasselti). Bateman slopes are relatively steep for male redbacks, and, as predicted by Bateman, there is little evidence for role reversal. Instead, males are competitive and show limited choosiness despite wide variation in female reproductive value. This study supports the proposal that high male mating investment coupled with low parental investment may predispose males to choosiness but will not lead to role reversal. We support the utility of using Bateman slopes to predict sex roles, even in systems with extreme male mating investment.