Sperm counts are declining in step with global increases in rates of obesity, suggesting the two factors may be linked. However, surprisingly, meta-analyses consistently fail to find evidence for a relationship between male body mass index (BMI) and semen quality. While BMI is the standard measure used to categorise obesity, not all obese (BMI > 30) individuals show signs of metabolic disease. Experiments using the Geometric Framework for Nutrition have revealed that mice fed a high carbohydrate/low fat/low protein diet become obese, yet do not develop impaired glucose tolerance. Hence, to test whether sperm function is related to body mass or metabolic health, we fed male mice one of 3 diets – high fat, high starch, or chow – for 12 weeks, and then assessed metabolic health and sperm quality. Both high fat and high starch males gained significantly more weight, driven by increases in fat mass, than chow-fed males. Yet, despite similar amounts of adiposity, high starch males had better glucose tolerance than high fat males. High starch males had smaller testes, but larger seminal vesicles, than either high fat or chow males. However, high starch males were able to compensate for their smaller testes, with no difference in sperm concentration found among diet treatments. Sperm motility and velocity was reduced in high fat, but not high starch males. The sperm of high starch males showed lower levels of oxidative stress than either high fat or chow males. No effect of diet treatment on sperm viability or DNA fragmentation was found. These results indicate that sperm quality is related to metabolic health, not fat mass. Hence, it appears that the lack of a correlation between BMI and male fertility is because obese, yet metabolically-healthy, males have normal sperm function.